Image detection method based on primary visual cortex cell characteristic

Abstract

The invention discloses an image detection method based on primary visual cortex cell characteristics; the method comprises the following steps: respectively building a space filter and a time filterso as to simulate a simple cell receptive field; integrating the space filter and the time filter so as to obtain a parity space-time filter; inputting an image sequence so as to respectively carry out convolution filtering treatment with the parity space-time filter, thus obtaining a simple cell response value sensitive to a specific direction [theta] and a specific rate Vc; integrating the simple cells so as to obtain complex cell responses; decoding the complex cell responses, building edge intensity functions, and realizing image edge detection. The built cell model can basically fit biological data, has direction selection characteristics, and the image detection method can solve the problems that a conventional edge detection method is poor in edge detection precision and large in anti-noise performance equalization difficulty.

Description

technical field [0001] The invention relates to an image processing technology inspired by biology, in particular to an edge detection method for optical images. Background technique [0002] As an important visual feature, edges contain image details and structural information. Image edge detection is one of the most basic technologies in the field of image processing, which can be applied to image retrieval, image analysis and computer vision. The biological vision system has incomparable advantages in image processing, and can effectively perceive and understand image information. Therefore, it has become an inevitable trend to use the research results in the field of biological vision to promote the development of computer vision. [0003] Most of the existing edge detection methods have a problem that the detection accuracy and anti-noise performance are difficult to balance. Although there are some methods that can solve this problem better, but greatly increase the a...

AI Technical Summary

Problems solved by technology

[0003] Most of the existing edge detection methods have a problem that the detection accuracy and anti-noise performance are difficult to balance. Although there are some methods that can solve this problem better, but greatly increase the amount of calculation, so it is easy to find a method with high detection performance. An edge detection method with good and strong anti-noise performance has always been an important research direction in the field of image processing.

Images