Research On The Application Of Improved Image Processing System Based On YOLO Model In Seed Purity Inspection

The purity of corn seed directly affects the quality and growth of corn,so the purity detection of corn seed has become an urgent demand in breeding and production.However,most of the traditional corn seed purity detection methods are based on manual detection,which has the problem of detection accuracy.Therefore,this paper proposes a corn seed purity detection method based on improved Yolo model.The content of this paper is as follows:(1)This paper first discusses the transformation relations among various color Spaces,introduces color features into plant classification,and applies the color feature vector of plant stem to seed classification research.(2)In terms of image texture feature extraction,this paper proposes a co-occurrence matrix model based on wavelet domain.Firstly,four filtered images of the original image are obtained by wavelet decomposition.Then four wavelet filter images are quantized to get four different images.Finally,each image is classified by extracting texture features using symbiosis matrix.(3)The identification model based on YOLO deep learning network is analyzed.Genetic algorithm and parameter selection improved particle swarm optimization network were used to train and identify seeds of different degrees.Through comparison and analysis,genetic algorithm and particle swarm optimization had their own advantages and disadvantages,so genetic algorithm was used to improve particle swarm optimization based on YOLO deep learning network.(4)The results obtained by biological test method and image recognition and processing system are compared to correct the results of image recognition and processing system,so that the final results of image recognition and processing system meet the requirements,and the accuracy of artificial identification results is as high as 99.5%.