Potatoes are one of the major food crops in the world. Potato black heart is a kind of defect that the surface is intact while the tissues in skin become black. This kind of potato has lost the edibleness, but it’s difficult to be detected with conventional methods. A nondestructive detection system based on the machine vision technology was proposed in this study to distinguish the normal and black heart of potatoes according to the different transmittance of them. The detection system was equipped with a monochrome CCD camera, LED light sources for transmitted illumination and a computer. Firstly, the transmission images of normal and black heart potatoes were taken by the detection system. Then the images were processed by algorithm written with VC++. As the transmitted light intensity was influenced by the radial dimension of the potato samples, the relationship between the grayscale value and the potato radial dimension was acquired by analyzing the grayscale value changing rule of the transmission image. Then proper judging condition was confirmed to distinguish the normal and black heart of potatoes after image preprocessing. The results showed that the nondestructive system built coupled with the processing methods was accessible for the detection of potato black heart at a considerable accuracy rate. The transmission detection technique based on machine vision is nondestructive and feasible to realize the detection of potato black heart.
Meat freshness is directly related to the health of consumers, and total volatile basic nitrogen (TVB-N) content is an important reference index for evaluating pork freshness. This paper attempted to measure TVB-N content for assessing pork meat freshness using a new self-developed portable and low cost detection device designed by ourselves basing on near infrared technique. The front-end part of this device was an integrated detection component containing a mini probe which was about 5cm in diameter circle. In the signal acquiring component, silicon photodiode detector was embedded in the center of light source in probe and spectral response range was 400-1100nm to receive diffuse light from pork meat surface in mini probe. The main circuits in this device included stabilized current supply circuit which was used to provide a stable power supply for each LED light source in probe and signal processing circuit which was utilized to complete signal amplification and A/D conversion, In addition, another vital function of the signal processing circuit was to analysis detection signals from mini probe in the detection component. For verifying this device performance, 58 pork samples with different freshness attributes and Multiple Linear Regression (MLR) mathematical method and ratio data processing algorithm were employed to build pork TVB-N content prediction model, and comparing with results from raw data model, the correlation coefficient of prediction and validation of TVB-N were 0.8027 and 0.7291 respectively, and the accuracy of predicting pork freshness was about 78.6%. This work demonstrates that it has the potential in nondestructive detection of TVB-N content in pork meat using this device, which can simplify related instruments design structure and reduce their development cost in future.