A nondestructive testing and automatic grading method for poultry egg surface cracks

A nondestructive testing and automatic grading method for poultry egg surface cracks comprises the steps that an image collecting device is used for obtaining an original image of an egg product, a vector median filtering method is used for smoothing the original image of the egg product, and image surface information and edge details are reserved; Graying the color image, and performing histogramequalization processing on the gray level image by using a histogram specifying method; Based on the image gray characteristic difference, obtaining optimal segmentation thresholds of the backgroundand the target through the inter-pixel variance to carry out image segmentation, and obtaining image contour information; Neural network training and learning are carried out by adopting an error backpropagation method, and nondestructive testing and automatic grading tasks of eggs are completed. The method has good practicability and stability, surface and edge details are reserved while image noise is removed through med