Aiming at the problem that the classical methods cannot meet the requirement of real-time detection and accuracy on far-infrared pedestrian detection,a far-infrared pedestrian detection method based on improved You Only Look Once (YOLO) model,which is a deep learning based model,is studied in this paper.By improving the input resolution of the deep convolutional neural network in YOLO and training the model on the real world far-infrared datasets,the detection model with the best performance is obtained.Besides,an adaptive resolution model based on the speed of the vehicle to give full play to the maximum performance of the detection system is proposed.Comparative experiments on the real world datasets show that compared with the traditional method,the accuracy rate of the proposed method increases from 76.5% to 89.2%,frames per second increases from 0.01259 f/s to 40.5 f/s,meeting the requirement of real-time detection.
参考文献
相似文献
引证文献
引用本文
谭康霞,平鹏,秦文虎.基于YOLO模型的红外图像行人检测方法[J].激光与红外,2018,48(11):1436~1442 TAN Kang-xia, PING Peng, QIN Wen-hu. Far-infrared pedestrian detection method based on YOLO model[J]. LASER & INFRARED,2018,48(11):1436~1442