THE DEVELOPMENT OF A MACHINE VISION SYSTEM FOR SHIITAKE GRADING
暂无分享,去创建一个
A grading system was developed to classify shiitake automatically into various grades of quality in terms of size, shape and color. A shiitake is scanned with a color charge-coupled-device camera and then various geometrical and color characteristics are determined by image processing. The online grading procedure includes the following sequences in processing dried shiitake: (a) color abnormality is eliminated by image analysis; (b) broken caps are detected by calculation of the area-to-perimeter ratio; and (c) size is determined by pixel counting from the bottom view. Image processing and algorithmic manipulations are done on a personal computer (PC) while grading is carried out by a programmable logic controller (PLC). The PLC accepts grading instructions from the PC, hence controls the grading mechanisms. Pilot system validation was done on 250 dried shiitake samples and resulted in a 97.6% accuracy rate. This result is better than that by visual inspection. A complete grading cycle takes 4.8 s, of which the PC unit participates 0.3‐ 0.7 s depending on the physical complexity of the shiitake being processed.
[1] Da-Wen Sun,et al. Inspection and grading of agricultural and food products by computer vision systems—a review , 2002 .
[2] C. T. Morrow,et al. Grading of Mushrooms Using a Machine Vision System , 1994 .
[3] Kalman Peleg. Quality Criteria of Sorting Operations , 1981 .
[4] Tünde Vı́zhányó,et al. Enhancing colour differences in images of diseased mushrooms , 2000 .