Image Recognition Sorting System

Aiming at the problem that it is difficult to realize the automatic feeding with orientation requirements for the parts with uniform appearance and shape and center of mass distribution in the automated production, the image acquisition, image processing, image feature extraction and classification recognition of the parts have been researched and analyzed. The image recognition as the core of the parts sorting system is proposed. A camera is installed at the discharge port of the vibrating plate for image acquisition, the wavelet transform is performed on the captured image to filter out the interference noise and dimensionality reduction, and the principal component analysis (PCA) is used to further reduce the dimensionality of the image and feature extraction, the extracted feature vectors are used as input vectors for the support vector machine (SVM), and the input vectors are used as samples through the SVM to identify the parts. The SVM recognizes the samples of the input vectors to determine the positional state of the parts, and finally the parts that do not meet the feeding requirements are pushed out of the feeding track by the driving device, so as to realize the automatic feeding for the next process.