Low level light(LLL) image contains rich information on environment details, but is easily affected by the weather. In the case of smoke, rain, cloud or fog, much target information will lose. Infrared image, which is from the radiation produced by the object itself, can be "active" to obtain the target information in the scene. However, the image contrast and resolution is bad, the ability of the acquisition of target details is very poor, and the imaging mode does not conform to the human visual habit. The fusion of LLL and infrared image can make up for the deficiency of each sensor and give play to the advantages of single sensor. At first, we show the hardware design of fusion circuit. Then, through the recognition probability calculation of the target(one person) and the background image(trees), we find that the trees detection probability of LLL image is higher than that of the infrared image, and the person detection probability of the infrared image is obviously higher than that of LLL image. The detection probability of fusion image for one person and trees is higher than that of single detector. Therefore, image fusion can significantly enlarge recognition probability and improve detection efficiency.
A method for moving target detection and segmentation using Markov random field (MRF)-based evaluation metric in infrared videos has been proposed. Starting with the most useful seeds of a moving object, which are extracted based on the “holes” effect of temporal difference; the proposed method employs a region growing method using local gray information and a spatial and temporal MRF model-based evaluation metric without ground truth for moving target segmentation in infrared videos. The segmented mask of a moving target is grown from the most useful seeds using the region growing method with thresholds. The proposed evaluation metric is utilized to determine the best growing threshold, where the performance of moving target segmentation is measured by that of segmented mask’s boundary. Thus, an MRF modeling for each boundary point of the segmented mask in spatial and temporal directions was considered by us. This problem is formulated using maximum a posteriori (MAP) estimation principle. At last, the global optimum of MRF-MAP framework is achieved using simulated annealing algorithm. The best segmented mask of a moving target is grown from the most useful seeds with the best growing threshold. Experimental results are reported to demonstrate the accuracy and robustness of our algorithm.
To research the attenuation performance of the AlGaN photocathode, three samples with same structures grown by metalorganic chemical vapor deposition (MOCVD) were activated with three different activation methods, which are Cs-only, Cs-O and Cs-O-Cs activation, respectively. The spectral responses and attenuated photocurrents of three AlGaN photocathodes were measured, the result shows that the Cs-O activated AlGaN photocathode have the lowest attenuation speed in the first few hours, the next are Cs-O-Cs and Cs-only activation, respectively. After the Cs-O-Cs activation sample has attenuated 90 min, its attenuation photocurrent curve is coincident with the Cs-O activation sample in the next measurement. The main factor which affects the photocurrent attenuation is Cs atom desorbed from the photocathodes surface.
Proc. SPIE. 9142, Selected Papers from Conferences of the Photoelectronic Technology Committee of the Chinese Society of Astronautics: Optical Imaging, Remote Sensing, and Laser-Matter Interaction 2013
To find out the best infrared and visible fusion system of fusion algorithm which has excellent target detection characteristics in different environment, we proposed a new fusion algorithm selective rule. We also defined new concepts: fusion algorithm coefficient and the equivalent transmissivity of system. Using local-target contrast, local-target articulation to calculate fusion algorithm coefficient, we can estimate the target detection performance of fusion system when it working in different air humidity environment. Also, we make use of infrared and visible fusion system designed by ourselves to verify this method. Besides fusion algorithm coefficient, we also use subjective evaluation to evaluate the target detection performance of fusion algorithm. At last, the best algorithm or the method which is most consistent with human visual in different conditions were found. Through this work, we can provide the basis for the algorithm of choice in the fusion system.
A new method for unsupervised segmentation of moving objects in infrared videos is presented. This method consists of two steps: difference image quantization and spatial segmentation. In the first step, the changed pixels in the difference image are quantized to several classes by using Bayes decision. It can be used to cluster the changed pixels belonging to the same moving object together. The pixels of the difference image are replaced by their corresponding class labels, thus forming a class-map of the difference image. In the second step, each class in the class-map is considered as a subset of the possible seeds of moving objects. A self-adaptive region growing method is then used to image segmentation on the basis of these different subsets. One of the focuses of this work is on spatial segmentation, where a criterion is proposed for evaluation of moving object segmentation without ground truth in infrared videos. This criterion is used to evaluate the performance of the segmentation masks grown from different subsets of the possible seeds. The best segmented image is determined to be the final segmentation result. Experiments show the advantage and robustness of the proposed algorithm on real infrared videos.
To improve the performance of GaAs NEA photocathodes, an exponential-doping structure GaAs material has been put forward, in which from the GaAs bulk-to-surface doping concentration is distributed exponentially from high to low. We apply this exponential-doping GaAs structure to the transmission-mode GaAs photocathodes. This sample was grown on the high quality
p-type Be-doped GaAs (100) substrate by MBE. We have calculated the band-bending energy in exponential-doping GaAs emission-layer, and the total band-bending energy is 59 meV which helps to improve the photoexcited electrons movement towards surface for the thin epilayer. The integrated sensitivity of the exponential-doping GaAs photocathode samples reaches 1547uA/lm.
To analyze the formation mechanism of the halo on low light level image intensifiers and the influencing factors on the halo size, a halo tester has been designed. Under the illumination between 10-2 lx and 10-4 lx, we use the tester to collect a 0.1922 mm hole image directly with CoolSNAPK4 charge-coupled device (CCD) in a darkroom. The practical measurement result shows that the amplification ratio is 343.4. Then we put the super second and third generation image intensifiers after the hole, and the halo sizes of the hole images on the screens are determined as 0.2388 and 0.5533 mm respectively. The results are helpful to improve the quality of the low light level image intensifiers.
The stability for reflection-mode GaN photocathode has been investigated by monitoring the photocurrent and the
spectral response at room temperature. We watch that the photocurrent of the cathode decays with time in the vacuum
system, and compare the spectral response curves after activation and after degradation. The photocurrent decay
mechanism for reflection-mode NEA GaN photocathode was studied by the surface model ［GaN (Mg) :Cs］:O-Cs. The
reduction of the effective dipole quantity, which is caused by harmful gases, is the key factor of the photocurrent
An evaluation for objectively assessing the quality of visible and infrared color fusion image is proposed. On the basis of the consideration that human perception is most sensitive to color, sharpness, and contrast when assessing the quality of color image, we propose four objective metrics: image sharpness metric (ISM), image contrast metric (ICM), color colorfulness metric (CCM), and color naturalness metric (CNM). The ISM is evaluated by image gradient information. The ICM is defined based on both gray and color histogram characteristics. A color chroma metric, as well as a color variety metric based on a color difference gradient, is proposed, respectively, to define the CCM. The CNM is defined by measuring the color distribution's similarity between the fusion image and nature image, which are of the same scene. All the color attributions are computed in the CIELAB color space. Experimental results show that the proposed objective metrics are meaningful and effective on color fusion image evaluation because they correspond well to subjective evaluation.
To improve the performance of GaAs NEA photocathodes, an exponential-doping structure GaAs material has
been put forward, in which from the GaAs bulk-to-surface doping concentration is distributed gradiently from high to
low. We apply this exponential-doping GaAs structure to the transmission-mode GaAs photocathodes. This sample was
grown on the high quality p-type GaAs (100) substrate by MBE with p-type Be doping. We have calculated the
band-bending energy in exponential-doping GaAs emission-layer, and the total band-bending energy is 59 meV which
helps improve the photoexcited electrons movement towards surface for the thin epilayer. The integrated sensitivity of
the two exponential-doping GaAs photocathode samples with different thickness reaches 1228uA/lm and 1547uA/lm
A fusion algorithm for infrared and visible light images based on region segmentation and the dual-tree complex wavelet
transform. Before image segmentation, morphological top-hat filtering is firstly performed on the IR image and visual
images respectively and the details of the luminous area are eliminated. Morphological bottom-hat filtering is then
performed on the two kinds of images respectively and the details of the dark area are eliminated. Make the top-hat
filtered image subtract the bottom-hat filtered image and obtain the enhanced images. Then the threshold method is used
to segment the enhanced images. After image segmentation, the DTCWT coefficients from different regions are merged
separately. Finally the fused image is obtained by performing inverse DTCWT. The evaluation results show the validity
of the presented algorithm.
A real-time color image fusion system has been presented for the infrared thermal camera and the low-light-level camera,
which provides more complete spectral image information. The statistical transform method based on the YCRCBcolor model
transfers the first order statistics of the color distribution of a representative natural color daytime reference image to the false
color dual -band images. This mapping is usually performed in a perceptually decorrelated color space. The colors in the
resulting colorized dual-band images closely resemble the colors in the daytime color reference image. Also, object colors
remain invariant under panning operations and are independent of the scene content. Preliminary field trials demonstrate the
potential of these systems for applications like surveillance, security and target detection.
The real-time processing system of infrared and Low level light(LLL) image fusion is developed. The system
consists of uncooled infrared imaging system, LLL TV system, real-time image processor, image acquisition card,
computer and monitor. Infrared imaging system is based on a 384×288-element uncooled microbolometer focal
plane arrays. LLL TV system uses a super Gen-llimage intensifier. The real-time image processor is designed
which can process the video outputs of uncooled infrared imaging system and LLL TV system. The real-time
image fusion based on weighted pixel average is relized by image processor. The image fusion algorithms based on
PCA weighted pixel average and pyramidal decomposition are simulated in computer. The results are given and
Fluorescence screen of Image intensifier is the key part to imaging quality of micro light and ultraviolet Image
intensifier. To research the performance testing and analysis of Fluorescence screen seems more important in China. The
research will help to understand the performance of Fluorescence screen, know where improvement should be made
and then a best performance entire tube will be achieved. This article will do the theory analysis to part of testing
instrument, area source electron gun's uniformity. Electron gun consists of taper tantalum filament, vacuum environment
and axial symmetry high pressure static field. The uniformity of hot electron emission of filament has been analyzed.
Upon that, this article will specially analyze the uniformity of electron in the effective area after they go through the axial
symmetry high pressure static field and get accelerated.
When testing the uniformity of Image intensifier fluorescence screen brightness, the million scale CCD brightness meter
is used. Due to the distance between the meter and fluorescence screen, the effect of ambient light on the testing result
is essential to the design of testing system. Test with super second generation tube, input a constant voltage to insure the
fluorescence screen brightness to be constant. Collect the brightness of the same fluorescence screen in different ambient
luminance environment of 1×102Lx, 1×101Lx, 1Lx, 1×10-1Lx, 1×10-2Lx, 1×10-3Lx. Study the results with software
MATLAB. It is concluded as: In ambient luminance environment of 1×10-1Lx the CCD has the best result. The
testing result in ambient luminance environment of above 1×103Lx show untrue image. The testing result in ambient
luminance environment of below 1×10-3Lx shows its own noise image and is unbelievable either.
Microbolometer focal plane array (FPA), as a popular kind of uncooled infrared detector, has a wide range of low cost
thermal imaging applications due to its high sensitivity and simple micro-fabrication process. The performance of
microbolometer imaging system is determined by many factors such as the property of the FPA, the effect of
nonuniformity correction, the condition of operation and so on. In this paper, the micro-structure and heat transfer
mechanism of microbolometer FPA are analysed to find out the substrate temperature characteristic. The response
nonuniformity of the FPA and corresponding two-point correction method are discussed to find out the calibration
temperature characteristic. And the power dissipation property of the thermal-electrical cooler (TEC) integrated under
the FPA is described to find out the ambient temperature characteristic. According to the simulation and experiment
results obtained from a 320×240 amorphous silicon microbolometer imaging system, it is concluded that all these
temperature parameters have a great influence on the system performance and should be well considered for different
working conditions to gain high system performance and imaging quality.
An uncooled thermal imaging system with multiple working temperatures will be presented. Transient response
performance of α-si microbolometer detectors is simulated firstly when the working temperature varies in the range from
-40deg. to +60deg. Simulating results show that α-si microbolometer detectors have coherent response performance in a
large range of working temperature, which lay basis for designing uncooled thermal imaging system with multiple
working temperatures. Different from traditional thermal imaging systems, this thermal imaging system has three
working temperature with an accuracy range of less than ±0.01deg. When working, the temperature control circuit will
switch between the working temperatures according to the variety of the environmental temperature. To evaluate this
thermal imaging system, we measure its power consumption and NETD in the environmental temperature range from
-40deg. to +60deg. The measurement results are that the total power is less than 2500mW and the NETD is less than
120mk. This indicates that the thermal imaging system has nearly the same imaging quality and obviously lower power,
compared with traditional thermal imaging systems.
An infrared target tracking system has been introduced in details for unmanned monitor application and its
corresponding method for extracting and tracking moving targets from real-time infrared video has also been described.
To ensure its real-time implementation on the tracking system, mature motion estimation techniques such as the
time-domain statistics method and the DT method are adopted in the tracking method which includes three stages: target
extracting, target classification and target tracking. A two-strategy classification method is adopted to improve
classification accuracy. The tracking process involves correlation matching between a template and the current motion
regions. The motion region with the best correlation is tracked and is used to update the template for subsequent tracking.
The infrared target tracking system is based on a high-speed DSP chip with an internet interface, which may transmit the
doubtful targets information to monitor center in time. To illustrate the effectivity of the infrared target tracking system,
experimental results has been given in the end of this paper.
Reasons that thermal imaging systems consume power have been analyzed, and a low-power design scheme of thermal
imaging systems has been presented with multiple working temperature points. Transient response performance of α-si
microbolometer detectors is simulated firstly when the working temperature varies in the range from -40°C to +60°C.
Simulating results show that α-si microbolometer detectors have coherent response performance in a large range of
working temperature, which lay basis for designing uncooled thermal imaging system with multiple working
temperatures. Different from traditional thermal imaging systems, this thermal imaging system has three working
temperature with an accuracy range of less than ±0.01°C. When working, the temperature control circuit will switch
between the working temperatures according to the variety of the environmental temperature. To evaluate this thermal
imaging system, we measure its power consumption and NETD in the environmental temperature range from -40°C to
+60°C. The measurement results are that the total power is less than 2500mW and the NETD is less than 120mk. This
indicates that the thermal imaging system has nearly the same imaging quality and obviously lower power, compared
with traditional thermal imaging systems.
Microbolometer detector is very competent as uncooled infrared detector for a wide range of thermal imaging
applications, since it has been found to be more sensitive and has the advantage of using standard Si micro-fabrication
process compared with pyroelectric or ferroelectric technology. The heart of microbolometer detector is a two
dimensional array of thermal sensitive thin-film layers, which can change their temperatures and resistivities depending
on the radiation absorbed. During the entire thermal imaging process, the microbolometer detector's substrate
temperature, calibration temperature and ambient temperature are the key parameters which determine the thermal-electrical
performance and the ultimate imaging quality of the microbolometer detector. In this work, based on the
analysis of the characteristics of these parameters, the experiment has been conducted with the uncooled infrared thermal
imaging system based on 320×240 amorphous silicon microbolometer detector working at different substrate
temperatures, adopting different calibration temperatures for different ambient temperatures. The corresponding
measurement results of the system's NETD, residual nonuniformity and power consumption, as well as the system's
imaging results are presented, which all have a great agreement of the theory analysis above.
Recent advances of microelectromechanical system (MEMS) technology have opened new opportunities for amorphous
silicon (α-Si) microbolometer focal plane arrays (FPAs) both for military and civil applications. α-Si membrane is
chosen for sensitive material of microbolometer FPAs due to its high temperature coefficient of resistance (TCR), high
resistivity and good mechanical properties. However, α-Si membrane also has the disadvantage of high 1/f noise, strict
preparation conditions and metastable effect. So nowadays, researches are focused on properties of α-Si membrane to
gain high performance of microbolometer FPAs. Since the pulsed bias readout mode of microbolometer FPAs causes a
non-steady-state of α-Si membrane during the operation, the transient thermal-electrical response process of the sensing
pixel is analyzed detailedly in this paper to predict the thermal and electrical performance of α-Si microbolemeter FPAs
such as responsivity, noise equivalent temperature difference (NETD), detectivity and power dissipation. Numerical
simulations are presented to investigate the factor which affects the performance of α-Si microbolometer FPAs. The
imaging experiment results obtained from a 320×240 α-Si microbolemeter FPA are in good agreement with the
theoretical analysis. The way to improve the performance of α-Si microbolemeter FPAs is given in the end of this paper.
Recent advances in MEMS and focal plane array (FPA) technologies have led to the development of manufacturing
microbolometers monolithically on a readout integrated circuit (ROIC). Since the response of microbolometer detectors
depends on the modification of temperature in micromachined bridge structures, it is useful to model and simulate
thermally the corresponding structures in order to predict their performance parameters. In this work, finite element
methods are performed to simulate the transient temperature field of thermistor films of microbolometer detectors. The
varisized supporting legs' impacts on the performance of detectors are discussed and the transient response for three
microbolometer configurations was investigated. At the same time, variation of the operation temperature's impacts on
total noise, noise equivalent to temperature difference (NETD) and detectivity (D*) are also discussed in details. These
performance analyses are helpful for optimum design of microbolometer infrared detectors' structure and rational choice
of operation temperature of infrared focal plane arrays.
Thermal imager can transfer difference of temperature to difference of electric signal level, so can be application to
medical treatment such as estimation of blood flow speed and vessel 1ocation, assess pain and so on. With the
technology of un-cooled focal plane array (UFPA) is grown up more and more, some simple medical function can be
completed with un-cooled thermal imager, for example, quick warning for fever heat with SARS. It is required that
performance of imaging is stabilization and spatial and temperature resolution is high enough. In all performance
parameters, noise equivalent temperature difference (NETD) is often used as the criterion of universal performance. 320
x 240 α-Si micro-bolometer UFPA has been applied widely presently for its steady performance and sensitive
responsibility. In this paper, NETD of UFPA and the relation between NETD and temperature are researched. several
vital parameters that can affect NETD are listed and an universal formula is presented. Last, the images from the kind
of thermal imager are analyzed based on the purpose of detection persons with fever heat. An applied thermal image
intensification method is introduced.
Since x-ray was discovered and applied to the imaging technology, the x-ray imaging techniques have experienced
several improvements, from film-screen, x-ray image intensifier, CR to DR. To store and transmit the image information
conveniently, the digital imaging is necessary for the imaging techniques in medicine and biology. Usually as the
intensifying screen technique as for concerned, to get the digital image signals, the CCD was lens coupled directly to the
screen, but which suffers from a loss of x-ray signal and resulted in the poor x-ray image perfonnance. Therefore, to
improve the image performance, we joined the brightness intensifier, which, was named the Low Light Level (LLL)
image intensifier in military affairs, between the intensifying screen and the CCD and designed the novel x-ray imaging
system. This design method improved the image performance of the whole system thus decreased the x-ray dose.
Comparison between two systems with and without the brightness intensifier was given in detail in this paper. Moreover,
the main noise source of the image produced by the novel system was analyzed, and in this paper, the original images
produced by the novel x-ray imaging system and the processed images were given respectively. It was clear that the
image performance was satisfied and the x-ray imaging system can be used in security checking and many other
nondestructive checking fields.
Uncooled microbolometer infrared detectors are being developed for a wide range of thermal imaging applications. To design and manufacture high-performance microbolometer infrared detectors, numerical calculation and simulation is necessary. In this work, finite element methods are performed to simulate the transient temperature field of thermistor films of microbolometer infrared detectors. The varisized supporting legs' impacts on the performance of detectors are discussed. At the same time, variation of the bias voltage and the substrate temperature's impacts on total noise, noise equivalent to temperature difference (NETD) and detectivity (D*) are also discussed in details. These performance analyses are helpful for optimum design of microbolometer infrared detectors' structure and rational choice of working temperature of infrared focal plane arrays.
In this paper, on the base of simple introduction of inner structure of 320×240 pixels UFPA in electronics and
calorifics, the relationship of NETD (noise equivalent temperature difference) and bias voltage are researched and
presented through the formulas about noise and NETD. The relation between NETD and four kinds of temperatures is
presented. Moreover the two bias voltages are adjusted to observe the changing of NETD. Some experiments on power
consumption and image quality of thermal imaging system is done, the result data is given. On the basis of the theory and
experiments, how to enhance the NETD performance of UFPA (Focal Plane Array) at much lower or higher than room
temperature is researched by analyzing experiment data. At last, the conclusion is summarized: in order to get the best
image and the lest power consumption, we should adjust these parameters to find the optimized configuration at different
The great achievements were achieved in the manufacturing of uncooled microbolometer infrared focal plane arrays(UFPA). By this technique infrared system can be made in the formation of small volume, light weight, low price and being portable. It promotes greatly the utilization of infrared system in many fields. The main disadvantage of UFPA is non-uniformity. Despite non-uniformity of UFPA has been greatly improved, non-uniformity still restricts the performance of uncooled infrared system. In this paper, the attention is focused on the technology and methods measuring the non-uniformity of UFPA. The system that can measure the non-uniformity of UFPA and evaluate the image processing algorithms is developed. The measurement system consists of blackbody, infrared optics, control units, processing circuit, high-speed A/D converter, computer and software. To obtain the output signals of UFPA, the drive circuit and control circuit of thermoelectric stabilizer(TEC) of UFPA are developed. In the drive circuit, the CPLD device is employed to insure a small size circuit. In the TEC circuit, a kind of highly integrated and cost-effective, high-effiency, switch-mode driver is used to insure temperature stability of 0.01°C. The system is used to measure non-uniformity of microbolometer detectors which are produced by ULIS company. It can also present the evaluation of algorithm. The results are given and analyzed.
Nonuniformity is a pressing problem particularly for uncooled focal-plane arrays imaging systems. Based on the theory of Least Mean Square (LMS), an algorithm is developed to compensate for the nonuniformity response in focal-plane arrays (FPA). The proposed algorithm is able to reduce various errors caused by noises, sampling, etc. A correction model of multi-order function is presented which has least-mean square error, and then by utilizing response values at several known temperature points the coefficients of this function are obtained via the theory of LMS fitting. Taking one pixel as example, the graph of the correction error is drawn. The ability of the algorithm is demonstrated by using simulated and real data. What’s more, the effects of this algorithm in nonuniformity correction are revealed as opposed to those of some traditional algorithm. To compensate response drift along time, a method based on estimating motion in continuous frames is presented.
A 320×240-uncooled-microbolometer-based signal processing circuit for infrared focal-plane arrays is presented, and the software designs of this circuit system are also discussed in details. This signal processing circuit comprises such devices as FPGA, D/A, A/D, SRAM, Flash, DSP, etc., among which, FPGA is the crucial part, which realizing the generation of drive signals for infrared focal-plane, nonuniformity correction, image enhancement and video composition. The device of DSP, mainly offering auxiliary functions, carries out communication with PC and loads data when power-up. The phase locked loops (PLL) is used to generate high-quality clocks with low phase dithering and multiple clocks are to used satisfy the demands of focal-plane arrays, A/D, D/A and FPGA. The alternate structure is used to read or write SRAM in order to avoid the contradiction between different modules. FIFO embedded in FPGA not only makes full use of the resources of FPGA but acts as the channel between different modules which have different-speed clocks. What's more, working conditions, working process, physical design and management of the circuit are discussed. In software designing, all the function modules realized by FPGA and DSP devices, which are mentioned in the previous part, are discussed explicitly. Particularly to the nonuniformity correction module, the pipeline structure is designed to improve the working frequency and the ability to realize more complex algorithm.