Image recognition method and device and computer readable storage medium

Abstract

The embodiment of the invention discloses an image recognition method and device and a computer readable storage medium. The efficiency and accuracy of image recognition can be improved. The method comprises the following steps: collecting a target image, and obtaining point cloud data of the target image; performing geometric shape detection on the point cloud data to obtain target geometric shape parameters; acquiring a direction line of the target geometrical shape on a display interface of the target image; based on the direction line and a preset detection threshold value, carrying out effectiveness detection on the target geometrical shape parameters; when the target geometrical shape parameter is detected to be valid, performing parameter iteration processing based on the target geometrical shape parameter and the point cloud data to obtain a target parameter; and based on the target parameter, performing plane expansion on the surface of the target geometric shape in the target image to obtain an identification result.

Description

technical field [0001] The present application relates to the field of image processing, and in particular to an image recognition method and device, and a computer-readable storage medium. Background technique [0002] At present, when performing target detection and target recognition on image information, when the image contains a complex background, there are many interference information, which is prone to misidentification or recognition failure. The existing technology usually adds special detection to the algorithm to improve the accuracy of recognition. Accuracy, but increasing the complexity of the algorithm will increase the computing cost, affect power consumption and real-time calculation, and the recognition efficiency is low; further, for the situation that there are other graphics with the same shape as the target of interest in the same image, now Some algorithms cannot distinguish the target of interest, and false detections will occur. Contents of the in...

AI Technical Summary

Problems solved by technology

[0002] At present, when performing target detection and target recognition on image information, when the image contains a complex background, there are many interference information, which is prone to misidentification or recognition failure. The existing technology usually adds special detection to the algorithm to improve the accuracy of recognition. Accuracy, but increasing the complexity of the algorithm will increase the computing cost, affect power consumption and real-time calculation, and the recognition efficiency is low; further, for the situation that there are other graphics with the same shape as the target of interest in the same image, now Some algorithms cannot distinguish the target of interest, and false detections will occur

Images