Study On Melon External Quality Automatic Grading Based On Computer Vision

The melon is one kind of fruits in the market, melon's exterior quality mainly involves size, shape, color and surface defect. Under the static condition, study on the technology of automatic gradation for melon using computer vision, the main research results were as follows:1. To the image underlying bed processing filter, the enhancement, the marginal check and the background division algorithm has conducted the feasibility study, obtained has suited the melon the pretreatment algorithm2. This forworded a new algorithm identify stem based on computting the ratio of chain code length and the Euclidea distance between each most large curvature in all continuous concave points group.3. Real-time fruit surface defect inspection and recognition is still a challenging subject due to its complexity. In order to enhance the precision and speed of defect segmentation, a novel defect segmentation algorithm was developed. The true defect moldy region and stem calyx region distinguished by using gray-level co-occurrence matrix and BP artificial neural network.4. Study on melon size,shape,color and surface defects exterior quality respectively. Proposed a surface defects grading system based on the method of counting pixel, proposed a shape system based on the method of the ratio between the distance and variance from edge points to center point, proposed a color grading system based on the method of Y histogram in CMYK color model, proposed a size grading system based on the melon's fruit diameter.5. A hard ware and software system for melon grading was developed. The hardware system can Perform image capture function, and the software can be used for melon grading by size,shape,color and surface defects.