A method for extracting low-level visual information from complex scenes

Abstract

The invention belongs to the field of scene visual cognition, and in particular relates to a method for extracting underlying visual information of complex scenes. In order to solve the underlying feature value extraction problem of complex scenes. The invention introduces an improved convolutional neural network structure, uses four kinds of convolution filters to form a multi-depth analysis set to perform image semantic segmentation on scene images; introduces feature convolution filters to filter and extract regional semantics of complex scenes, and then extracts the results The transposed convolution is used to segment the semantic region of the scene image; the region segmentation result of the scene semantics is used as the activation bias and substituted into the final underlying visual information feature value extraction network, which can ensure that various types of scene details will not be lost. After the semantic segmentation of the scene area, the present invention can well extract the feature value of the underlying information of the scene required for the cognitive experiment, and can better preserve the details in the complex scene.

Description

technical field [0001] The invention belongs to the field of scene visual cognition, and in particular relates to a method for extracting underlying visual information of complex scenes. The present invention is mainly aimed at complex visual cognition experimental scenes through algorithmic realization, and extracts three types of dominant underlying visual attention resources in the global scene. Background technique [0002] Human's underlying visual processing mechanism and neuroscience research on visual cells point out that when people observe the scene without the influence of existing concepts, human beings tend to allocate visual attention resources to areas with high color saturation in the scene, Areas of high color contrast and areas with edge / directional features. These three types of attention resource content are called overt underlying visual features. In the neural signals of human visual attention, these three types of features occupy most of the informati...

AI Technical Summary

Problems solved by technology

[0005] In view of the defect that the existing low-level visual information eigenvalue method loses the details of the eigenvalue during the extraction, this defect will cause the visual cognition experiment of complex scenes to fail to obtain effective low-level visual eigenvalues ​​during scene analysis, and the lack of this part will be extremely serious. The Reliability of High Impact Cognitive Analysis

Images