Image Understanding Method Based on Synergy Fluctuation Force

Abstract

The invention discloses an image comprehension method based on synergetics fluctuation force, which uses synergetics to solve related problems in image comprehension. The image comprehension method based on synergetics fluctuation force in the present invention effectively utilizes prior knowledge by introducing potential function dynamic equation fluctuation force, perfects side inhibition item and adds a semantic ambiguity attention parameter item to solve multi-objective problems and ambiguity Solve related problems with specific goals, reflect the mutual guidance and mutual driving relationship between the scene and the goals in the scene, and make the system become "Everyone Takes Own Site", instead of only one winner in the "Winner Takes All" system, complete Scene Interpretation and Object Recognition in Scenes.

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