Infrared Image Vehicle Detection Based on Haar-like Feature
Danqi Chen, Guodong Jin, Libin Lu, Lining Tan, Wenle Wei · 2018
An infrared image vehicle detection algorithm based on Haar-like features is proposed in this paper. Firstly, the top-hat transformation and bottom-hat transformation are used to enhance the infrared vehicle image, the contrast between the vehicle and the background is improved. Secondly, we propose to use Haar-like features to describe and calculate infrared vehicles' features. Thirdly, the improved maximum entropy segmentation algorithm is used to segment the vehicle. After that, using the vehicle's prior knowledge to remove the detected false alarm target. Finally, the Fl-measure model is used to evaluate the algorithm. The experimental results show that the algorithm can effectively detect the vehicle targets in the infrared images taken under different conditions, and can partially meet the requirements of real-time detection and meet the accuracy requirements.