Polluted soils analysis and characterization is one of the basic step to perform in order to collect all the information to design and set-up correct soil reclamation strategies. Soil analysis is usually performed through "in-situ" sampling and laboratory analysis. Such an approach is usually quite expensive and does not allow to reach a direct and detailed knowledge of large areas for the intrinsic limits (high costs) linked to direct sampling and polluting elements detection. As a consequence numerical strategies are applied to extrapolate, starting from a discrete set of data, that is those related to collected samples, information about the contamination level of areas not directly interested by physical sampling. These models are usually very difficult to handle both for the intrinsic variability characterizing the media (soils) and for the high level of interactions between polluting agents, soil characteristics (organic matter content, size class distribution of the inorganic fraction, composition, etc.) and environmental conditions (temperature, humidity, presence of vegetation, human activities, etc.). Aim of this study, starting from previous researches addressed to evaluate the potentialities of hyperspectral imaging approach in polluting soil characterization, was to evaluate the results obtainable in the investigation of an "ad hoc" polluted benthonic clay, usually utilized in rubbish dump, in order to define fast and reliable control strategies addressed to monitor the status of such a material in terms of insulation.

Soil erosion and its related runoff is a serious problem in U.S. agriculture. USDA has classified 27% of U.S. agricultural land as being highly erodible. Because of the erosion, rivers, lakes, and water table are contaminated due to the agriculture chemicals such as nitrogen, phosphorus, and pesticides contained in the runoff water. This is a serious environmental problem nationwide. It is well recognized that residue coverage on the soil surface can reduce soil erosion. The objective of this paper was to explore the potential of using ASTER data for soybean plant residue cover estimation. In the spring of 2004, personnel from Natural Resource Conservation Service (NRCS) and Institute for Technology Development (ITD) did a traditional windshield survey in three Indiana Counties, Wabash, Huntington, and Grant. Fields with greater than 30% residue cover were classified as conservation tillage (no till); those with 16-30% residue cover as reduced tillage; and those with less than 15% residue cover as traditional tillage. ASTER data was collected over the study sites on April 14, 2004. Spectral information was extracted from the ASTER image for statistical analysis. Field values for various indices were calculated from the reflectance data. Residue coverage estimation from the survey was used as the ground truth for the field. Analysis was performed to determine the capability of ASTER data to identify crop residue coverage. The initial results indicated that ASTER imagery has moderate capability to identify residue coverage - or tillage practice within the soybean fields.

This study is aimed to develop a multi-functional remote sensing system based on spectral imaging and environmental sensing for seedling production in the greenhouses. The spectral images were grabbed with exposure time and signal gain controls through IEEE-1394 interface; and a color camera and a B/W camera with optical filter at 780 nm were used. A control program was developed to grab the good quality images using the automatic exposure algorithm with a developed software using Matlab and LabVIEW. To obtain necessary spectral information regarding tray locations and seedling growth status on greenhouse benches, a serial image processing procedures, including spatial calibration, image stitching, gray-level calibration and image segmentation were developed. The data of tray positions and growth status were transferred to the look up table (LUT) and delivered to the water management module through the DataSocket server and wireless network. Besides, the environmental sensing sub-system, including temperature, relative humidity, and lighting measurements, was also developed with the PCI-6023 interface to analyze the spatial distribution of these parameters in the greenhouse. The information of environmental status will provide a better management for seedling growth in greenhouses.

Biological cells have components acting as electrical elements that maintain the health of the cell by regulation of the electrical charge content. Plant impedance is decided by the state of plant physiology and pathology. Plant physiology and pathology can be studies by measuring plant impedance. The effect of Cucumber Mosaic Virus red bean isolate (CMV-RB) on electrical resistance of tomato leaves was studied by the method of impedance measurement. It was found that the value of resistance of tomato leaves infected with CMV-RB was smaller than that in sound plant leaves. This decrease of impedances in leaf tissue was occurred with increased severity of disease. The decrease of resistance of tomato leaves infected with CMV-RB could be detected by electrical resistance detecting within 4 days after inoculation even though significant visible differences between the control and the infected plants were not noted, so that the technique for measurement of tomato leaf tissue impedance is a rapid, clever, simple method on diagnosis of plant disease.

Bacterial contamination by Listeria monocytogenes puts the public at risk and is also costly for the food-processing industry. Traditional methods for pathogen identification require complicated sample preparation for reliable results. Previously, we have reported development of a noninvasive optical forward-scattering system for rapid identification of Listeria colonies grown on solid surfaces. The presented system included application of computer-vision and patternrecognition techniques to classify scatter pattern formed by bacterial colonies irradiated with laser light. This report shows an extension of the proposed method. A new scatterometer equipped with a high-resolution CCD chip and application of two additional sets of image features for classification allow for higher accuracy and lower error rates. Features based on Zernike moments are supplemented by Tchebichef moments, and Haralick texture descriptors in the new version of the algorithm. Fisher's criterion has been used for feature selection to decrease the training time of machine learning systems. An algorithm based on support vector machines was used for classification of patterns. Low error rates determined by cross-validation, reproducibility of the measurements, and robustness of the system prove that the proposed technology can be implemented in automated devices for detection and classification of pathogenic bacteria.

We developed an antibody-based fiber-optic biosensor to rapidly detect low levels of Escherichia coli O157:H7 and shiga-like toxins (SLTs) in ground beef samples. The principle of the sensor is a sandwich immunoassay using an antibody which is specific for E. coli O157:H7 or toxins. A polyclonal antibody was first immobilized on polystyrene fiber waveguides through a biotin-streptavidin reaction that served as the bacteria and toxin capture entity. Alexa Fluor 647 dye-labeled antibodies against E. coli O157:H7 or SLTS incubated with the waveguides were used to detect cells or toxin and generate a specific fluorescent signal, which was acquired by launching a 635 nm laser-light from an Analyte-2000. Fluorescent molecules within several hundred nanometers of the fiber were excited by an evanescent wave, and a portion of the emission light from fluorescent dye transmitted by the fiber and collected by a photodetector at wavelengths of 670 to 710 nm quantitatively. This immunosensor was specific for E. coli O157:H7 compared with multiple other foodborne bacteria. The approach was also able to detect ~0.5 μg/mL of pure SLTs and the the SLTs associated with 10 5 E. coli O157:H7 cells at stationary phase after olfoxacin induction.

We have developed a detection system and associated protocol based on optical forward scattering where the bacterial colonies of various species and strains growing on solid nutrient surfaces produced unique scatter signatures. The aim of the present investigation was to develop a bio-physical model for the relevant phenomena. In particular, we considered time-varying macroscopic morphological properties of the growing colonies and modeled the scattering using scalar diffraction theory. For the present work we performed detailed studies with three species of Listeria; L. innocua, L. monocytogenes, and L. ivanovii. The baseline experiments involved cultures grown on brain heart infusion (BHI) agar and the scatter images were captured every six hours for an incubation period of 42 hours. The morphologies of the colonies were studied by phase contrast microscopy, including measurement of the diameter of the colony. Growth curves, represented by colony diameter as a function of time, were compared with the time-evolution of scattering signatures. Similar studies were carried out with L. monocytogenes grown on different substrates. Non-dimensionalizing incubation time in terms of the time to reach stationary phase was effective in reducing the dimensionality of the model. Bio-physical properties of the colony such as diameter, bacteria density variation, surface curvature/profile, and transmission coefficient are important parameters in predicting the features of the forward scattering signatures. These parameters are included in a baseline model that treats the colony as a concentric structure with radial variations in phase modulation. In some cases azimuthal variations and random phase inclusions were included as well. The end result is a protocol (growth media, incubation time and conditions) that produces reproducible and distinguishable scatter patterns for a variety of harmful food borne pathogens in a short period of time. Further, the bio-physical model we developed is very effective in predicting the dominant features of the scattering signatures required by the identification process and will be effective for informing further improvements in the instrumentation.

To evaluate the applicability of non-invasive visible and near infrared (VIS-NIR) spectroscopy for determining ethanol concentration of Chinese rice wine in square brown glass bottle, transmission spectra of 100 bottled Chinese rice wine samples were collected in the spectral range of 350-1200 nm. Statistical equations were established between the reference data and VIS-NIR spectra by partial least squares (PLS) regression method. Performance of three kinds of mathematical treatment of spectra (original spectra, first derivative spectra and second derivative spectra) were also discussed. The PLS models of original spectra turned out better results, with higher correlation coefficient in calibration (R_cal) of 0.89, lower root mean standard error of calibration (RMSEC) of 0.165, and lower root mean standard error of cross validation (RMSECV) of 0.179. Using original spectra, PLS models for ethanol concentration prediction were developed. The R_cal and the correlation coefficient in validation (R_val) were 0.928 and 0.875, respectively; and the RMSEC and the root mean standard error of validation (RMSEP) were 0.135 (%, v v^-1) and 0.177 (%, v v^-1), respectively. The results demonstrated that VIS-NIR spectroscopy could be used to predict ethanol concentration in bottled Chinese rice wine.

The feasibility of non-invasive visible and near infrared (VIS-NIR) spectroscopy for determining wine age (1, 2, 3, 4, and 5 years) of Chinese rice wine was investigated. Samples of Chinese rice wine were analyzed in 600 mL square brown glass bottles with side length of approximately 64 mm at room temperature. VIS-NIR spectra of 100 bottled Chinese rice wine samples were collected in transmission mode in the wavelength range of 350-1200 nm by a fiber spectrometer system. Discriminant models were developed based on discriminant analysis (DA) together with raw, first and second derivative spectra. The concentration of alcoholic degree, total acid, and °Brix was determined to validate the NIR results. The calibration result for raw spectra was better than that for first and second derivative spectra. The percentage of samples correctly classified for raw spectra was 98%. For 1-, 2-, and 3-year-old sample groups, the sample were all correctly classified, and for 4- and 5-year-old sample groups, the percentage of samples correctly classified was 92.9%, respectively. In validation analysis, the percentage of samples correctly classified was 100%. The results demonstrated that VIS-NIR spectroscopic technique could be used as a non-invasive, rapid and reliable method for predicting wine age of bottled Chinese rice wine.

The current study describes the use of three forms of optical measurement of single wheat kernels for screening of Fusarium head blight for eventual incorporation in high-speed optical sorters. Our previous research has demonstrated a sorting efficiency of approximately 50 percent with existing high-speed equipment, but a much higher efficiency (~95%) when analytical spectrometers are used. The intention of the current work is to bridge this efficiency gap. Knowledge gained from analysis of the single kernel in-flight response will provide design criteria for improvement of high-speed optical sorters for recognition of mold-damaged wheat.

Aflatoxin contaminated corn poses a serious threat to both domestic animals and humans, because of its carcinogenic properties. Traditionally, corn kernels have been examined for evidence of bright greenish-yellow fluorescence (BGYF), which is an indication of possible presence of Aspergillus flavus, one of the aflatoxin producing strains of fungi, when illuminated with a high-intensity ultra-violet light. The BGYF test is typically the first step that leads to an in-depth chemical analysis for possible aflatoxin contamination. The objective of the present study was to analyze hyperspectral BGYF response of corn kernels under UVA excitation. The target corn samples were collected from a commercial corn field in 2005 and showed abundant BGYF response. The BGYF positive kernels were manually picked out and imaged under a visible near-infrared hyperspectral imaging system under UV radiation with excitation wavelength centered at 365 nm. Initial results exhibited strong emission spectra with peaks centered from 500 nm to 515 nm wavelength range for BGYF positive kernels. Aflatoxin levels on the BGYF positive and negative corn kernels (used as control) were measured subsequently with high performance liquid chromatography. The mean aflatoxin concentration level was 5114 ppb for the BGYF positive and undetectable for the normal kernels.

We recently found that optical scattering coefficients of beef muscle can be used as a predictor for beef tenderness. However, it is still not clear what specific muscle properties are responsible for optical scattering. As an effort to answer these questions, we conducted several controlled experiments in which we studied the changes of scattering coefficient with muscle sarcomere length and ageing time. The optical scattering coefficient of beef muscles were measured based on a diffusive fitting of spatially resolved reflectance measurements. Samples with different sarcomere lengths were obtained by proper carcass hanging strategies. Our results indicated that muscle scattering coefficient increased with ageing time and sarcomere length. These experimental observations can be qualitatively explained based on previous research on single muscle scattering. This study suggests that muscle structural properties have significant impact on muscle optical scattering coefficients.

Animal meat products may not be the best choice for many people in the world due to various reasons such as cost, health problems, or religious restrictions. High moisture (40-80%) extrusion technology shows a great promise for texturizing vegetable proteins into fibrous meat alternatives. Soy protein which is healthy, highly nutritious, low in both fat and carbohydrate has been used in high moisture extrusion process to produce meat analogs with well formed fiber that resemble chicken or turkey breast meat. Assessing fiber formation in extruded products is important for controlling extrusion quality in manufacturing process. Although several methods have been studied for quantifying fiber formation in extrudates, their applications for real time quality control in manufacturing process have been challenging. We explored the possibility of applying a nondestructive method based on backscattered reflectance to measure the fiber formation of extruded soy proteins. An image processing method was developed to extract the light reflectance profile at the extrudates' surface. We applied the anisotropic continuous time random walk (CTRW) theory to quantitatively describe the fiber formation in extrudates based on extracted surface reflectance profiles. This method has a potential to be used as a non-destructive, fast, real time quality control tool for products with fibrous structures.

Our recent studies indicated that the optical scattering has the ability to characterize the micro structures in beef muscles. In this study, optical scattering spectra of beef semimembranosus and longissimus muscle samples were measured at different cooking temperatures along with the corresponding Warner-Bratzler shear force values. Overall, scattering coefficients increased first and then decreased. The increase was likely due to the denaturation of myosin and the thermal-stable collagen cross links depending on the temperature ranges; while the decrease was attributed to the gelatinization of thermal-labile collagen cross links. The collagen content and the relative proportionality of the thermal-stable and the thermal-labile collagen cross links could affect the relative changes of the scattering coefficients at different temperatures. Our results indicated that the optical scattering can indeed reflect the states of the micro structures in beef muscles and have the potential to be used as an indicator for beef quality prediction.

This paper is concerned with the detection of bone fragments embedded in de-boned skinless chicken breast fillets by modeling images made by back-lighting and embedded bone fragments. Imaging of chicken fillets is often dominated by strongly multiple scattering properties of the fillets. Thus, resulting images from multiple scattering are diffused, scattered and low contrast. In this study, both transmittance and reflectance hyperspectral imaging, which is a nonionized and non-destructive imaging modality, is investigated as an alternative method to the conventional transmittance X-ray imaging technique which is an ionizing imaging modality. As a way of reducing the influence of light scattering on images and thus increasing the image contrast, the use of a structured line light is examined along with an image formation model that separates undesirable lighting effects from an image. The image formation model based on an illumination-transmittance model is applied for correcting non-uniform illumination effects so that embedded bones are more easily detected by a global threshold. An automated image processing algorithm to detect bones is also proposed. Experimental results with chicken breast fillets and bone fragments are provided.

Images of packaged raw chicken purchased in neighborhood supermarkets were captured via a digital camera in laboratory and home settings. Each image contained the surface reflectivity information of the chicken tissue. The camera's red, green and blue light signals fluctuated and each spectral signal exhibited a random series across the surface. The Higuchi method, where the length of each increment in time (or spatial) lag is plotted against the lag, was used to explore the fractal property of the random series. (Higuchi, T., "Approach to an irregular time series on the basis of fractal theory", Physica D, vol 31, 277-283, 1988). The fractal calculation algorithm was calibrated with the Weierstrass function. The standard deviation and fractal dimension were shown to correlate with the time duration that a package was left at room temperature within a 24-hour period. Comparison to packaged beef results suggested that the time dependence could be due microbial spoilage. The fractal dimension results in this study were consistent with those obtained from yeast cell, mammalian cell and bacterial cell studies. This analysis method can be used to detect the re-refrigeration of a "left-out" package of chicken. The extension to public health issues such as consumer shopping is also discussed.

Sliced ham products undergo significant discolouration and fading when placed in retail display cabinets. This is due to factors such as illumination of the display cabinet, packaging, i.e. low OTR (Oxygen Transmission Rate) or very low OTR packaging, product to headspace ratio and percentage of residual oxygen. This paper presents initial investigations into the development of a sensor to measure rate of colour fading in cured ham, in order to predict an optimum colour sell-by-date. An investigation has been carried out that shows that spectral reflections offer more reproducibility than CIE L*a*b* readings, which are, at present, most often used to measure meat colour. Self-Organising Maps were then used to classify the data into five colour fading stages, from very pink to grey. The results presented here show that this classifier could prove an effective system for determining the rate of colour fading in ham.

Near infrared (NIR) spectroscopy is an ideal analytical method for rapid and nondestrctive measurement of the properties of agriculture products. The efficient use of this method is dependent on multivariate calibration methods determined by the sensitivity to variations. However, fluctuation of background and noise are unavoidable during collecting spectra, which will not only worsen the precision of prediction, but also complicate the multivariate models. Therefore, the first step of a multivariate calibration based on NIR spectra data is often to preprocess the data for the purpose of removing the varying background and noise. In this study, wavelet transform (WT) was used to eliminate the varying background and noise simultaneously in the near infrared spectroscopy signals of 55 navel oranges. Three families of mother wavelets (Symlets, Daubechies and Coiflet), four threshold selection rules (Rigrsure, Heursure, Minimaxi, Fixed form threshold), and two threshold functions (soft and hard) were applied to estimate the performances. The sugar content of intact navel orange was calculated by partial least squares regression (PLSR) with the reconstructed spectra after denoised. The results show that the best denoising performance was reached via the combination of Daubechies 5, "Fixed form" threshold selection rule, and hard threshold function. Based on the optimization parameter, wavelet regression models on sugar content in navel orange were also developed and resulted in a smaller prediction error than a traditional PLSR model.

Fourier transform near infrared reflectance (FT-NIR) spectroscopy has been used successfully to measure soluble solids content (SSC) in citrus fruit. However, for practical implementation, the technique needs to be able to compensate for fruit temperature fluctuations, as it was observed that the sample temperature affects the near infrared reflectance spectrum in a non-linear way. Temperature fluctuations may occur in practice because of varying weather conditions or improper conditioning of the fruit immediately after harvest. Two techniques were found well suited to control the accuracy of the calibration models for soluble solids with respect to temperature fluctuations. The first, and most practical one, consisted of developing a global robust calibration model to cover the temperature range expected in the future. The second method involved the development of a range of temperature dedicated calibration models. The drawback of the latter approach is that the required data collection is very large. The global temperature calibration model avoids temperature-sensitive wavelengths for the calibration of SSC. Global temperature models are preferred above dedicated temperature models because of the following shortcomings of the latter. For each temperature, a new calibration model has to be made, which is time-consuming.

Watermelon is a popular fruit in the world. Soluble solids content (SSC) is major characteristic used for assessing watermelon internal quality. This study was about a method for nondestructive internal quality detection of watermelons by means of visible/Near Infrared (Vis/NIR) diffuse transmittance technique. Vis/NIR transmittance spectra of intact watermelons were acquired using a low-cost commercially available spectrometer when the watermelon was in motion (1.4m/s) and in static state. Spectra data were analyzed by partial least squares (PLS) method. The influences of different data preprocessing and spectra treatments were also investigated. Performance of different models was assessed in terms of root mean square errors of calibration (RMSEC), root mean square errors of prediction (RMSEP) and correlation coefficient (r) between the predicted and measured parameter values. Results showed that spectra data preprocessing influenced the performance of the calibration models and the PLS method can provide good results. The nondestructive Vis/NIR measurements provided good estimates of SSC index of watermelon both in motion and in static state, and the predicted values were highly correlated with destructively measured values. The results indicated the feasibility of Vis/NIR diffuse transmittance spectral analysis for predicting watermelon internal quality in a nondestructive way.

Fluorescence and reflectance (or interactance) are promising techniques for measuring fruit quality and condition. Our previous research showed that a hyperspectral imaging technique integrating fluorescence and reflectance could improve predictions of selected quality parameters compared to single sensing techniques. The objective of this research was to use a low cost spectrometer for rapid acquisition of fluorescence and interactance spectra from apples and develop an algorithm integrating the two types of data for predicting skin and flesh color, fruit firmness, starch index, soluble solids content, and titratable acid. Experiments were performed to measure UV light induced transient fluorescence and interactance spectra from 'Golden Delicious' apples that were harvested over a period of four weeks during the 2005 harvest season. Standard destructive tests were performed to measure maturity parameters from the apples. Principal component (PC) analysis was applied to the interactance and fluorescence data. A back-propagation feedforward neural network with the inputs of PC data was used to predict individual maturity parameters. Interactance mode was consistently better than fluorescence mode in predicting the maturity parameters. Integrating interactance and fluorescence improved predictions of all parameters except flesh chroma; values of the correlation coefficient for firmness, soluble solids content, starch index, and skin and flesh hue were 0.77, 0.77, 0.89, 0.99, and 0.96 respectively, with the corresponding standard errors of 6.93 N, 0.90%, 0.97 g/L, 0.013 rad, and 0.013 rad. These results represented 4.1% to 23.5% improvements in terms of standard error, in comparison with the better results from the two single sensing methods. Integrating interactance and fluorescence can better assess apple maturity and quality.

Development of nondestructive measurements of soluble solids and firmness, which are two important ripeness and quality attributes of fruits, benefits the producers, processors and packers. The objective of this research was to evaluate the use of near-infrared (NIR) spectroscopy in detecting soluble solid content (SSC) and firmness for pears of three cultivars 'Cuiguan', 'Xueqing' and 'Xizilv' (n=160 of each cultivar). Relationships between nondestructive NIR spectral measurements and firmness and SSC of pear fruits were established by partial least square regression (PLSR) method. Models were developed for each cultivar, every two cultivars, and for all three cultivars in the spectral range of 800-2500 nm. The results of the models for all three cultivars turned out the best. For SSC assessment: correlation coefficients of calibration (r_cal), root mean standard errors of calibration (RMSEC) and root mean standard errors of prediction (RMSEP) were 0.93, 0.35 °Brix and 0.50 °Brix for all three cultivars, respectively. For firmness assessment: rcal, RMSEC and RMSEP were0.92, 2.29 N, 2.95 N for all three cultivars, respectively. The results indicate that NIR spectroscopy can be used for predicting SSC and firmness of pear fruit and are the basis for the development of NIR analyzer suitable for on line application.

The Gradient Vector Flow (GVF) snake was used for color fruit shape detection, which is proposed by Chengxiang Xu, this snake has two well properties than traditional snake: large capture range and its ability to move into boundary concavities. Indicator and morphological operation before applying GVF snake firstly preprocess the color fruit image. In our experiments, we compared the detection result of this approach to traditional snake and traditional edge operators and it is obvious that the performance of this approach is better; the boundaries detected by GVF snake are thin and smooth, which are very important for fruit size detection and shape classification.

We have developed nondestructive opto-electronic imaging techniques for rapid assessment of safety and wholesomeness of foods. A recently developed fast hyperspectral line-scan imaging system integrated with a commercial apple-sorting machine was evaluated for rapid detection of animal feces matter on apples. Apples obtained from a local orchard were artificially contaminated with cow feces. For the online trial, hyperspectral images with 60 spectral channels, reflectance in the visible to near infrared regions and fluorescence emissions with UV-A excitation, were acquired from apples moving at a processing sorting-line speed of three apples per second. Reflectance and fluorescence imaging required a passive light source, and each method used independent continuous wave (CW) light sources. In this paper, integration of the hyperspectral imaging system with the commercial applesorting machine and preliminary results for detection of fecal contamination on apples, mainly based on the fluorescence method, are presented.

Development of machine vision systems to examine fruit for quality and contamination problems has been stalled due to lack of an inexpensive, fast, method for appropriately orienting fruit for imaging. We recently discovered that apples could be oriented based-on inertial properties. Apples were rolled down a ramp consisting of two parallel rails. When sufficient angular velocity was achieved, the apples moved to a configuration where the stem/calyx axis was perpendicular to the direction of travel. This discovery provides a potential basis for development of a commercially-viable orientation system. However, many question remain concerning the underlying dynamic principles that govern this phenomenon. An imaging system and software were constructed to allow detailed observation of the orientation process. Sequential 640×480 monochrome images are acquired at 60 fps and 1/500 sec exposure. The software finds the center of the apple in each image as well as the vertical movement of the track at a selected coordinate. Early tests revealed that the compliance of the track played a significant role in the orientation process. These data will be used to compare results from empirical tests with predictions of dynamic models.

Automatic diagnosis of plant disease is important for plant management and environmental preservation in the future. The objective of this study is to use multispectral reflectance measurements to make an early discrimination between the healthy and infected plants by the strain of tobacco mosaic virus (TMV-U1) infection. There were reflectance changes in the visible (VIS) and near infrared spectroscopy (NIR) between the healthy and infected plants. Discriminant models were developed using discriminant partial least squares (DPLS) and Mahalanobis distance (MD). The DPLS models had a root mean square error of calibration (RMSEC) of 0.397 and correlation coefficient (r) of 0.59 and the MD model correctly classified 86.7% healthy plants and up to 91.7% infected plants.

Detection of skin tumors on chicken carcasses is considered. A chicken skin tumor consists of an ulcerous lesion region surrounded by a region of thickened-skin. We use a new adaptive branch-and-bound (ABB) feature selection algorithm to choose only a few useful wavebands from hyperspectral data for use in a real-time multispectral camera. The ABB algorithm selects an optimal feature subset and is shown to be much faster than any other versions of the branch and bound algorithm. We found that the spectral responses of the lesion and the thickened-skin regions of tumors are considerably different; thus we train our feature selection algorithm to separately detect the lesion regions and thickened-skin regions of tumors. We then fuse the two HS detection results of lesion and thickened-skin regions to reduce false alarms. Initial results on six hyperspectral cubes show that our method gives an excellent tumor detection rate and a low false alarm rate.

In the U. S. egg industry, anywhere from 130 million to over one billion infertile eggs are incubated each year. Some of these infertile eggs explode in the hatching cabinet and can potentially spread molds or bacteria to all the eggs in the cabinet. A method to detect the embryo development of incubated eggs was developed. Twelve brown-shell hatching eggs from two replicates (n=24) were incubated and imaged to identify embryo development. A hyperspectral imaging system was used to collect transmission images from 420 to 840 nm of brown-shell eggs positioned with the air cell vertical and normal to the camera lens. Raw transmission images from about 400 to 900 nm were collected for every egg on days 0, 1, 2, and 3 of incubation. A total of 96 images were collected and eggs were broken out on day 6 to determine fertility. After breakout, all eggs were found to be fertile. Therefore, this paper presents results for egg embryo development, not fertility. The original hyperspectral data and spectral means for each egg were both used to create embryo development models. With the hyperspectral data range reduced to about 500 to 700 nm, a minimum noise fraction transformation was used, along with a Mahalanobis Distance classification model, to predict development. Days 2 and 3 were all correctly classified (100%), while day 0 and day 1 were classified at 95.8% and 91.7%, respectively. Alternatively, the mean spectra from each egg were used to develop a partial least squares regression (PLSR) model. First, a PLSR model was developed with all eggs and all days. The data were multiplicative scatter corrected, spectrally smoothed, and the wavelength range was reduced to 539 - 770 nm. With a one-out cross validation, all eggs for all days were correctly classified (100%). Second, a PLSR model was developed with data from day 0 and day 3, and the model was validated with data from day 1 and 2. For day 1, 22 of 24 eggs were correctly classified (91.7%) and for day 2, all eggs were correctly classified (100%). Although the results are based on relatively small sample sizes, they are encouraging. However, larger sample sizes, from multiple flocks, will be needed to fully validate and verify these models. Additionally, future experiments must also include non-fertile eggs so the fertile / non-fertile effect can be determined.

Several of visible and NIR bands were sought to explore the potential for the classification of fecal / ingesta ("F/I") objectives from rubber belt and stainless steel ("RB/SS") backgrounds. Spectral features of "F/I" objectives and "RB/SS" backgrounds showed large differences in both visible and NIR regions, due to the diversity of their chemical compositions. Such spectral distinctions formed the basis on which to develop simple three-band ratio algorithms for the classification analysis. Meanwhile, score-score plots from principal component analysis (PCA) indicated the obvious cluster separation between "F/I" objectives and "RB/SS" backgrounds, but the corresponding loadings did not show any specific wavelengths for developing effective algorithms. Furthermore, 2-class soft independent modeling of class analogy (SIMCA) models were developed to compare the correct classifications with those from the ratio algorithms. Results indicated that using ratio algorithms in the visible or NIR region could separate "F/I" objectives from "RB/SS" backgrounds with a success rate of over 97%.

During in-plant testing of a hyperspectral line-scan imaging system, images were acquired of wholesome and systemically diseased chickens on a commercial processing line moving at a speed 70 birds per minute. A fuzzy logic based algorithm using four key wavelengths, 468 nm, 501 nm, 582 nm, 629 nm, was developed using image data from the validation set of images of 543 wholesome and 66 systemically diseased chickens. A classification method using the fuzzy logic based algorithm was then tested on the testing set of images of 457 wholesome and 37 systemically diseased chickens, as well as 80 systemically diseased chickens that were imaged off-shift during breaks between normal processing shifts of the chicken plant. The classification method correctly identified 89.7% of wholesome chicken images and 98.5% of systemically diseased chicken images in the validation set. For the testing data set, the method correctly classified 96.7 % of 457 wholesome chicken images and 100% of 37 systemically diseased chicken images. The 80 images acquired off-shift were also 100% correctly identified.

A prototype real-time multispectral imaging system for fecal and ingesta contaminant detection on broiler carcasses has been developed. The prototype system includes a common aperture camera with three optical trim filters (517, 565 and 802-nm wavelength), which were selected by visible/NIR spectroscopy and validated by a hyperspectral imaging system with decision tree algorithm. The on-line testing results showed that the multispectral imaging technique can be used effectively for detecting feces (from duodenum, ceca, and colon) and ingesta on the surface of poultry carcasses with a processing speed of 140 birds per minute. This paper demonstrated both multispectral imaging hardware and real-time image processing software. For the software development, the Unified Modeling Language (UML) design approach was used for on-line application. The UML models included class, object, activity, sequence, and collaboration diagram. User interface model included seventeen inputs and six outputs. A window based real-time image processing software composed of eleven components, which represented class, architecture, and activity. Both hardware and software for a real-time fecal detection were tested at the pilot-scale poultry processing plant. The run-time of the software including online calibration was fast enough to inspect carcasses on-line with an industry requirement. Based on the preliminary test at the pilot-scale processing line, the system was able to acquire poultry images in real-time. According to the test results, the imaging system is reliable for the harsh environments and UML based image processing software is flexible and easy to be updated when additional parameters are needed for in-plant trials.

The objective of this research is to design and fabricate a compact, cost effective multispectral instrument and to collect and analyze spectra for real-time contaminant detection for poultry processing plants. It was revealed by our previous research that the fecal contamination on the surface of the poultry carcass could be detected by sensing the spectral reflectance of the carcass surface in two specific wavelengths, namely 517 nm and 565 nm. The prototype system developed in this research consists of a multispectral imaging system, illumination system and handheld PC. To develop the system cost-effectively, all components are selected from off-the-shelf products and manually assembled. The multispectral imaging sensor developed in this research is a two-port imaging system that consists of two identical monochrome cameras, optical system and two narrow bandpass filters whose center of the wavelength are 520 and 560 nm, respectively. A spectral reflectance from a chicken carcass is collected and split in two directions by an optical system including a beamsplitter and lenses, and then two identical collimated lights are filtered by the narrow bandpass filters and delivered to the cameras. Lens distortions and geometric misalignment of the two cameras are mathematically compensated to register two images perfectly. The prototype system is tested in the real environment and shows that it can effectively detect feces and ingesta on the surface of poultry carcasses.

An online line-scan imaging system was developed for differentiation of wholesome and systemically diseased chickens. The hyperspectral imaging system used in this research can be directly converted to multispectral operation and would provide the ideal implementation of essential features for data-efficient high-speed multispectral classification algorithms. The imaging system consisted of an electron-multiplying charge-coupled-device (EMCCD) camera and an imaging spectrograph for line-scan images. The system scanned the surfaces of chicken carcasses on an eviscerating line at a poultry processing plant in December 2005. A method was created to recognize birds entering and exiting the field of view, and to locate a Region of Interest on the chicken images from which useful spectra were extracted for analysis. From analysis of the difference spectra between wholesome and systemically diseased chickens, four wavelengths of 468 nm, 501 nm, 582 nm and 629 nm were selected as key wavelengths for differentiation. The method of locating the Region of Interest will also have practical application in multispectral operation of the line-scan imaging system for online chicken inspection. This line-scan imaging system makes possible the implementation of multispectral inspection using the key wavelengths determined in this study with minimal software adaptations and without the need for cross-system calibration.

The objective of this research is to develop a digital image analysis algorithm for detection of multiple rice seeds images. The rice seeds used for this study involved a hybrid rice seed variety. Images of multiple rice seeds were acquired with a machine vision system for quality inspection of bulk rice seeds, which is designed to inspect rice seeds on a rotating disk with a CCD camera. Combining morphological operations and parallel processing gave improvements in accuracy, and a reduction in computation time. Using image features selected based on classification ability; a highly acceptable defects classification was achieved when the algorithm was implemented for all the samples to test the adaptability.

By using imaging techniques, plant physiological parameters can be assessed without contact with the plant and in a non-destructive way. During plant-pathogen infection, the physiological state of the infected tissue is altered, such as changes in photosynthesis, transpiration, stomatal conductance, accumulation of Salicylic acid (SA) and even cell death. In this study, the different temperature distribution between the leaves infected by tobacco mosaic virus strain-TMV-U1 and the noninfected leaves was visualized by digital infrared thermal imaging with the microscopic observations of the different structure within different species tomatoes. Results show a presymptomatic decrease in leaf temperature about 0.5-1.3 °C lower than the healthy leaves. The temperature difference allowed the discrimination between the infected and healthy leaves before the appearance of visible necrosis on leaves.

A combined laser 3D and X-ray imaging system is newly developed for food safety inspection. Two kinds of cameras are used in this system. One is CCD camera which is used to provide an accurate thickness profile of the object and the other is X-ray line-scan camera which is to get the high resolution X-ray image. A unique three-step calibration procedure is proposed to calibrate these two kinds of cameras. Firstly, the CCD camera is calibrated to link the CCD pixels to points in 3D world coordinate system. Secondly, the X-ray line-scan camera is calibrated to link points in 3D world coordinate system to the X-ray line sensors. The X-ray fan beam effect is also compensated in this stage. Finally, direct mapping from CCD pixel to X-ray line sensor is realized using the information from the first two calibration steps. Based on the calibration results, look-up tables are also generated to replace the expensive runtime computation with simpler lookup operation. Results show that high accuracy has been achieved with the whole system calibration.

In this study, two measuring systems for chlorophyll content of tomato leaves were developed based on near-infrared spectral techniques. The systems mainly consists of a FT-IR spectrum analyzer, optic fiber diffuses reflection accessories and data card. Diffuse reflectance of intact tomato leaves was measured by an optics fiber optic fiber diffuses reflection accessory and a smart diffuses reflection accessory. Calibration models were developed from spectral and constituent measurements. 90 samples served as the calibration sets and 30 samples served as the validation sets. Partial least squares (PLS) and principal component regression (PCR) technique were used to develop the prediction models by different data preprocessing. The best model for chlorophyll content had a high correlation efficient of 0.9348 and a low standard error of prediction RMSEP of 4.79 when we select full range (12500-4000 cm^-1), MSC path length correction method by the log(1/R). The results of this study suggest that FT-NIR method can be feasible to detect chlorophyll content of tomato leaves rapidly and nondestructively.

A machine vision system for real-time fruit quality inspection was developed. The system consists of a chamber, a laser projector, a TMS-7DSP CCD camera (PULNIX Inc.), and a computer. A Meteor-II/MC frame grabber (Matrox Graphics Inc.) was inserted into the slot of the computer to grab fruit images. The laser projector and the camera were mounted at the ceiling of the chamber. An apple was put in the chamber, the spot of the laser projector was projected on the surface of the fruit, and an image was grabbed. 2 breed of apples was test, Each apple was imaged twice, one was imaged for the normal surface, and the other for the defect. The red component of the images was used to get the feature of the defect and the sound surface of the fruits. The average value, STD value and comentropy Value of red component of the laser scatter image were analyzed. The Standard Deviation value of red component of normal is more suitable to separate the defect surface from sound surface for the ShuijinFuji apples, but for bintang apples, there is more work need to do to separate the different surface with laser scatter image.

A machine vision system for egg weight detection was developed. Egg image was grabbed by a CCD camera and a frame grabber. An indicator composed of R, G, B intensity was used for image segmentation. A series of algorithms were developed to evaluate egg's vertical diameter, maximal horizontal diameter, upper horizontal diameter and nether horizontal diameter. Based on extracted four size features of vertical and maximal/upper/nether horizontal diameter, a regression model between egg's weight and its size was established using SAS, which was used to detect egg's weight. The experiment results indicated that, for egg weight detection on the machine vision system, the correlative coefficient of the regression model was 0.9781 and the absolute error was no more than ±3 g, which would be lower work load on human graders and an increased flexibility in the egg quality control process in egg's industrialization.

Sugar content (SC) is very important factors of navel orange internal quality and can be measured non-destructive by visible and near infrared spectroscopy. The feasibility of visible and near infrared spectroscopy for nondestructively measuring SC of navel orange fresh juices was investigated by means of spectral transmittance technique. A total 55 juice samples were used to develop the calibration and prediction models. Three different kinds of mathematical treatments (original, first derivative and second derivative) of spectra in the range of 400-800 nm were discussed and two kinds of reference standards were used. Different spectra correction algorithms (constant, multiplicative signal correction (MSC) and standard normal variate (SNV) were compared. Three kinds of calibration models including partial least square (PLS) regression, stepwise multiple linear regression (SMLR) and principle component regression (PCR) were evaluated for the determination of SC in navel orange juice. Performance of different models was assessed in terms of root mean square errors of prediction (RMSEP) and correlation coefficient (r) of validation set of samples. The correlation coefficients of calibration models for SC was 0.97, the correlation coefficients of prediction models for SC was 0.86, and the corresponding RMSEP was 0.56. The results show that visible near infrared transmittance technique is a feasible method for non-destructive measurement of sugar content of fruit juice.

There is increase pressure to reduce the use of pesticides in modern crop production to decrease the environment impact of current practice and to lower production costs. It is therefore imperative that sprays are only applied when and where needed. However it is difficult to measure the severity of plant disease as a result of the irregular leaf and disease spots shapes. In this research, a pixel method is proposed, and the severity of plant disease was graded accuracy by using technology of image analysis, and then the method was compared with traditional method for measured of plant infection severity. The leaves images were acquired by a CCD camera and transferred to a host computer and were stored as files in TIFF format. From the experimental results, it shows that the image method has an acceptable accuracy; and image processing is a rapid and non-destructive way to gain the plant infection severity.

The potential of using Near Infrared diffuse reflectance spectroscopy to assess soluble solids content (SSC) of intact navel orange was examined. A total 40 samples were used to develop the calibration and prediction models. NIR spectral data were collected in the spectral region between 350 nm and 2500 nm and its second derivative spectra was used for this study. Different scattering correction algorithms (no preprocessing and multiplicative scattering correction (MSC) were compared. Calibration models based on different spectral ranges, different derivatives and different kinds of statistical models including partial least square (PLS) and principle component regression (PCR) were also compared in this research. The best results of PLS models with the second derivative spectra are r=0.929, RMSEC=0.517 and RMSEP=0.592, in the wavelength range of 361-2488 nm. The segment length used to derivate the spectra influences the calibration model and the results are better when the segment lengths and gap sizes are lower in Norris derivate filter. The results show that this method is feasible for rapid assessing SSC of the navel orange.

Visible/near infrared spectroscopy on-line determination had been widely used in agricultural products and food samples non-destructive internal quality determination. This research proposed to design real-time determination software in order to estimate soluble solids content (SSC) of fruit on line. Functions of the software included real-time spectroscopy pre-processing, real-time spectroscopy viewing, model building, SSC estimating, etc. In addition, Fenghua juicy peaches were used to validate the practicability and the real-time capability. And SSCs of peach samples were predicted by the software on line. The research provided some help to the real-time non-destructive internal quality determination of the fruit. As the important part of the real-time determination, the determination method and technology were fully accordance with the need at real-time and model's precision.

Spectral performance would be affected by many factors such as temperature, equipment parameters and so on. Humidity fluctuations may occur in practice because of varying weather conditions. The objective of this research was to find out whether the change of humidity would influence the near infrared spectrum of samples. In this trial, an airproof, humidity-controllable test-bed was established to change the humidity of the mini environment. At 40%, 50%, 60%, 70% and 80% degrees of humidity, each sample's final spectrum was attained by removing the background's spectrum from the sample's. For whether the influence of the sample's and the background's spectrum are equal was not known, This trial was divided into two groups: detecting background and sample at each degree of humidity (group 1) and background's detecting just happened at 40% degree of humidity (group 2). This research was based on the hardware of NEXUS intelligent FT-IR spectrometer, made by Nicolet instrument company U.S.A, with using fiber optic diffuse reflectance accessory. The final spectrum was analysed using single variance analysis and Mahalanobis Distance methods. The result shows that neither in group 1 nor 2, humidity had little influence on NIR.

In this research, a method to identify Kuler fragrant pear's sexuality with machine vision was developed. Kuler fragrant pear has male pear and female pear. They have an obvious difference in favor. To detect the sexuality of Kuler fragrant pear, images of fragrant pear were acquired by CCD color camera. Before feature extraction, some preprocessing is conducted on the acquired images to remove noise and unnecessary contents. Color feature, perimeter feature and area feature of fragrant pear bottom image were extracted by digital image processing technique. And the fragrant pear sexuality was determined by complexity obtained from perimeter and area. In this research, using 128 Kurle fragrant pears as samples, good recognition rate between the male pear and the female pear was obtained for Kurle pear's sexuality detection (82.8%). Result shows this method could detect male pear and female pear with a good accuracy.