Image understanding method based on synergetics fluctuating force

Abstract

The invention discloses an image understanding method based on synergetics fluctuating force, which solves relevant problems in image understanding by utilizing the synergetics. The image understanding method based on the synergetics fluctuating force of the invention solves the relevant problems by inducting the potential function dynamical equation fluctuating force to effectively utilize prior knowledge, perfecting lateral inhibition items and adding a semantic polysemy attention parameter item to solve the problems of multiple targets and ambiguity targets, therefore, the relationship of mutual guidance and mutual driving between a scene and a target in the scene is reflected, and the system is changed into 'Everyone Takes Own Site' instead of 'Winner Takes All' only with one winner, thus finishing scene interpretation and target identification in the scene.

Description

technical field [0001] The invention relates to the fields of image understanding, computer vision and synergetics, in particular to an image understanding method based on synergistic fluctuation force. Background technique [0002] In recent years, computer vision technology has developed rapidly. Scene description and object recognition in the scene are the core tasks in image understanding. Let the computer recognize and judge what objects are in the scene and where they are, so as to solve the "what-where" problem. [0003] In image understanding, scene interpretation and target recognition in the scene are the key points, and many researchers are devoted to the research of scene interpretation and target recognition in the scene. The traditional scene interpretation method mainly has two parts. One is to focus on the research of visual psychology and physiology, mainly to study the psychological and physiological mechanism of rapid scene perception; the other is to stud...

AI Technical Summary

Problems solved by technology

The existing image understanding methods based on this have the following problems: how to use the scene information to guide the selection of the target; how to construct the prior information of the scene to guide the target recognition in the scene; Prior Information for Object Recognition and Scene Understanding in Scenes

[0005] In short, the existing image understanding methods are difficult to use prior knowledge to improve the accuracy of understanding. In complex scenes, it is impossible to learn, analyze and understand multiple targets at the same time, and it is difficult to effectively judge the ambiguity and polysemy of the target. Deficiencies in decision-making, etc.

Images