首页>
外国专利>
MULTI-SCALE SENSING PEDESTRIAN DETECTION METHOD BASED ON IMPROVED FULL CONVOLUTIONAL NETWORK
MULTI-SCALE SENSING PEDESTRIAN DETECTION METHOD BASED ON IMPROVED FULL CONVOLUTIONAL NETWORK
展开▼
机译：基于改进的全卷积网络的多尺度传感行人检测方法
展开▼
页面导航
摘要
著录项
相似文献
摘要
The present invention belongs to the field of pedestrian detection, and relates to a multi-scale sensing pedestrian detection method based on an improved full convolutional network. The method comprises: first, importing a deformable convolutional layer in a fully convolutional network structure to expand a receptive field of a feature map; then, extracting a multi-scale pedestrian suggestion region by means of a cascaded RPN, importing a multi-scale determination strategy, defining a scale determination layer, and determining a scale category of the pedestrian suggestion region; and finally constructing a multi-scale sensing network, importing a Soft-NMS detection algorithm, and combining a classification value and a regression value of each network output to obtain a final pedestrian detection result. Experiments show that the detection algorithm of the present invention produces lower detection errors on benchmark pedestrian detection data sets Caltech and ETH, which is superior to the accuracy of all detection algorithms in the current data set, and is suitable for detecting a pedestrian on a far scale.
展开▼
