Far-infrared pedestrian detection method based on distinctive features

Abstract

The invention relates to a far-infrared pedestrian detection method based on distinctive features. According to the invention, for an acquired far-infrared image, two distinctive features are found out, wherein two distinctive features comprise a highlight feature corresponding to a human head and a torso perpendicular feature corresponding to a human body; two regions are found out and fused to acquire an interested region; a gradient direction histogram feature is extracted from the interested region; a trained linear support vector machine is used to carry out pedestrian detection; at different temperatures, a pedestrian region can be accurately and efficiently extracted; fast and robust pedestrian detection is realized; and the driving safety factor of a driver at night, in a harsh environment and in the case of low visibility is improved.

Description

technical field [0001] The invention belongs to the field of image processing technology and pattern recognition technology, and in particular relates to a far-infrared pedestrian detection method based on salient features. Background technique [0002] The increase in car ownership has to some extent led to an increase in car accidents. With the frequent occurrence of traffic accidents, the issue of car driving safety is becoming more and more important. In traffic accidents, the main victims are pedestrians and cyclists in road traffic. For driving safety, many researchers at home and abroad have conducted research on pedestrian detection systems based on visible light, and achieved certain results. [0003] However, pedestrian detection systems based on visible light usually perform poorly in dark complex and harsh environments. Nighttime traffic accidents are mostly caused by unclear vision at night, unsatisfactory lighting conditions, and too short vision of the driv...

AI Technical Summary

Problems solved by technology

Patent (3) adopts the method of mixed Gaussian model for modeling, mainly for pedestrian detection in static background, the effect of pedestrian detection in moving scenes is not good, and there are certain requirements for computing performance

[0012] Therefore, although the existing patents can detect pedestrians in some situations, they generally have poor adaptability and low robustness, and cannot maintain a good effect in various environments.

Especially when some ambient temperatures are high, the accuracy is low and the real-time performance cannot meet the requirements

Images