An image recognition method, apparatus, device and computer-readable storage medium

Abstract

Embodiments of the present invention disclose an image recognition method, apparatus, device, and computer-readable storage medium. Among them, the method includes performing gray image description processing based on a structure balance network on the image to be identified, and obtaining an initial structure balance matrix of the image to be identified; taking the initial structure balance matrix and a pre-built changed structure balance matrix as a Hadamard product to perform the sum of the structure balance matrix. The target structure balance matrix is ​​obtained by the change between the target structure balance matrix; the characteristic parameters of the target structure balance matrix are extracted to be used for target recognition of the image to be recognized, and the characteristic parameters are the average positive connection strength and negative connection strength average value of the target structure balance matrix. The technical solution of the present application does not consider the topology of the network, simplifies the complexity of the network model, reduces the amount of computation, and enhances the noise tolerance of the recognition method, thereby improving the speed and recognition accuracy of image recognition; the gray value is used as the structure balance in the network. The connection strength is improved, the discrimination of the recognition target is improved, and the recognition rate is improved.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of machine vision, and in particular, relate to an image recognition method, device, equipment, and computer-readable storage medium. Background technique [0002] With the rapid development of artificial intelligence technology, machine vision technology, as an important branch of artificial intelligence development, has also been developed accordingly. Machine vision technology integrates technologies in optics, machinery, electronics, computer software and hardware, etc. , pattern recognition, artificial intelligence, signal processing, optomechanical integration and other fields. Image recognition technology (recognizing target objects in images) is an important part of machine vision technology. The accuracy and speed of target recognition are important indicators for measuring image recognition. Improving the accuracy and speed of target recognition is the key to improving image r...

AI Technical Summary

Problems solved by technology

[0005] However, when the contour shape of the image changes with the influence of light intensity or occlusion and other factors, the relevance and order of the points in the image contour point set will be destroyed. If the image contour is regarded as a series of sequential continuous points set, it will increase the difficulty of extracting the precise contour shape, and the topology between the threshold and the gray level difference needs to be considered as the recognition parameters, resulting in low recognition accuracy and recognition efficiency of the entire image recognition process

Images