Multi-channel Detection System For Pesticide Residues Based On Image Analysis

In recent years, food poisoning often occurs because of eating food which contains pesticide residues. If people eat food with excessive pesticide residues for a long time, it can cause chronic poisoning, carcinogenic. Teratogenic may also increase.More and more people have paid attention to food safety issues, so the development of rapid and efficient pesticide residue detection system is important for food safety.The multi-channel detection system for pesticide residues proposed in this paper is based on the principle of enzyme inhibition. Under normal circumstances,acetylcholine is hydrolyzed under the catalytic action of cholinesterase. The reaction of hydrolyzate and reagent will generate a yellow solution. But if sample solution to be detected contains pesticide residues, the pesticide residues will inhibit the enzyme activity. Catalysis decrease will make the color of the solution faded. Pesticide residues can be judged by the color similarity of the sample solution and the standard solution.The system which includes 96 channels sample cell and combines with image analysis can detect a large number of samples of pesticide residues quickly and efficiently at the same time. Through the man-machine interface, users can operate easily. At the same the pesticide detection results will be displayed on the screen and the results also can be saved.The system software is completed under the VC ++ 6.0 development platform,combined with Open CV. It includes image acquisition, image display, image preprocessing, 96 channels segmentation, color extraction and determination and so on. Image acquisition is accomplished through the secondary development of SDK provided by USB200 image acquisition card. Image display is achieved by programming with function provided by Open CV and the MFC controls. In the image pre-processing, first, a copy of the acquired image is created, and then gray the image;second, canny edge detection algorithm is used to extract the edge, which contains theGaussian filter. Image segmentation is realized by Hough transform circle detection technology in edge binary image. Two circles are separated within the specified pixel range to obtain the center coordinates and the radius. Then all channels can be found by the radius and the distance between center coordinates. Color extraction is completed in the square area within the circle. The side length of the square is half of the radius. Then the R,G,B values of each pixel are obtained and the average values of R,G,B are considered to be the color information of this channel. Color similarity judgment is to use color information of sample solution and standard solution, to calculate the saturation similarity coefficient, brightness similarity coefficient. Then the color similarity coefficient can be obtained. And then color similarity decision rule is used to determine the similarity of color about sample solution and standard solution. Pesticide judgment means that if the color of the sample solution and standard solution are not similar, the sample solution contains pesticide residues,positive.In order to verify the effect of the detection system, the present system was used to detect the sample solution. And the results were compared with the test results of spectrophotometer. The experimental results show that the system can detect a large number of samples at a faster speed and lower cost while ensuring the accuracy.