Two-dimensional contour shape recognition and classification method based on local geometrical characteristic sequence modeling

Abstract

The invention discloses a two-dimensional contour shape recognition and classification method based on local geometrical characteristic sequence modeling in the technical field of object recognition and classification, and solves the problems that existing manual adjustment and setting of shape features are too complex and the success rate of object two-dimensional contour shape recognition and classification is low, and the method is characterized by comprising the steps of extracting local geometrical features of a target object shape; converting the extracted feature information into time sequence information through LSTM; modeling the local geometrical characteristic sequence based on a recurrent neural network; performing feature selection and balance based on a full-connection neuralnetwork to acquire a classification result of a target object. The defect that shape features are manually adjusted and set to be too complex is overcome. The success rate of object two-dimensional contour shape recognition and classification is remarkably increased. Two-dimensional contour shapes of different objects can be distinguished, and fine shape differences including fine local differences are captured.

Description

technical field [0001] The invention relates to the technical field of object recognition and classification, more specifically, it relates to a two-dimensional contour shape recognition and classification method based on local geometric feature sequence modeling. Background technique [0002] The machine vision system converts the captured target into an image signal through the machine vision product (that is, the image capture device, which is divided into CMOS and CCD), and transmits it to a dedicated image processing system to obtain the shape information of the captured target. According to the pixel distribution and Information such as brightness and color is converted into digital signals; the image system performs various operations on these signals to extract the characteristics of the target, and then controls the on-site equipment actions according to the results of the discrimination. In many application fields of machine vision, the recognition and classificati...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a two-dimensional contour shape recognition and classification method based on local geometric feature sequence modeling, which has the feature of adding object shape features to the deep learning network, reflecting the relationship between object shape features, that is, overcoming artificial adjustment settings The shortcoming of the shape features being too complex has significantly improved the success rate of object two-dimensional contour shape recognition and classification, and can distinguish the two-dimensional contour shapes of different objects and capture fine shape differences, including subtle local differences

Images