Machine vision-based mouse exploration identification method

Abstract

The invention discloses a machine vision-based mouse exploration identification method. The method comprises the steps of acquiring an image by utilizing an existing surveillance camera; denoising a mouse contour in the image by utilizing a background segmentation technology; and after extraction, performing comprehensive intelligent identification. A behavior of enabling the nose of a mouse to keep a certain range (which can be defined as needed) away from a target object and face the passing target object is regarded as an exploration behavior, and exploration time is recorded. According to the method, various machine vision technologies are integrated and include technologies of Canny edge detection, corner point detection, centroid detection and the like, and a proportion constraint-based corner point detection algorithm and a centroid detection-based corner point prediction algorithm are proposed, so that nose exploration behavior parameters of an experimental mouse can be accurately identified and the accuracy reaches up to 97%; and the method does not need to be artificially supervised, so that the manpower and the economic cost are greatly reduced and the problem of unreal and non-rigorous achievements caused by artificial subjective judgments in scientific experiments is effectively solved.

Description

technical field [0001] The invention relates to the field of image analysis and recognition, in particular to a method for monitoring mouse image changes and mouse behavior recognition by using machine vision. Background technique [0002] Behavioral analysis of experimental animals is an important technical means for the study of advanced nervous system functions. It is not only widely used in traditional fields such as neuropsychology and neuropharmacology, but also extends to emerging interdisciplinary research fields such as artificial intelligence, bionics, and brain-computer interface. [0003] The behavior of animals is composed of posture and posture changes, kinematic parameters are only a rough description of animal behavior, and animal posture contains more psychological information, which has more important significance in behavior analysis. [0004] Early animal behavior monitoring was mainly based on direct observation and manual recording, and there were many...

AI Technical Summary

Problems solved by technology

[0004] Early animal behavior monitoring was mainly based on direct observation and manual recording, and there were many problems such as qualitative subjectivity, time-consuming and labor-intensive.

Images