首页> 外国专利> METHOD AND DEVICE FOR DETECTING VEHICLE OCCUPANCY USING PASSENGER'S KEYPOINT DETECTED THROUGH IMAGE ANALYSIS FOR HUMANS' STATUS RECOGNITION

METHOD AND DEVICE FOR DETECTING VEHICLE OCCUPANCY USING PASSENGER'S KEYPOINT DETECTED THROUGH IMAGE ANALYSIS FOR HUMANS' STATUS RECOGNITION

机译:利用图像识别分析通过乘客关键点检测车辆占用率的方法和装置

摘要

A method for detecting vehicle occupancy using a passenger keypoint based on analysis of a vehicle''s interior image, the method comprising: (a) when the interior image is acquired, the vehicle occupancy detection device, (i) features the interior image By inputting into the extraction network, a convolution operation is applied to the indoor image to generate a feature tensor, and (ii) a keypoint heatmap and a PAF are generated by inputting the feature tensor into a keypoint heatmap & part affinity field (PAF) extractor. (Iii) by inputting the keypoint heatmap and the PAF to the keypoint detection device, extracting keypoints from the keypoint heatmap, and (iv) grouping the keypoints based on the PAF, each Detecting a key point for each passenger; And (b) by inputting the key point to a seat occupancy matcher, matching the passenger to the seat by referring to the input key point and a preset region of interest (ROI) of the seat, thereby matching the vehicle. And detecting the occupancy.
机译:一种基于乘客对车辆内部图像的分析使用乘客关键点检测车辆占用率的方法,该方法包括:(a)当获取内部图像时,车辆占用率检测设备,(i)将内部图像特征化为输入到提取网络中,对室内图像进行卷积运算以生成特征张量,并且(ii)通过将特征张量输入到关键点热图和部分亲和力字段(PAF)提取器中来生成关键点热图和PAF 。 (ii)通过将关键点热图和PAF输入到关键点检测设备,从关键点热图中提取关键点,并且(iv)基于PAF对关键点进行分组,每个针对每个乘客检测关键点;并且(b)通过将关键点输入到座位占用匹配器,通过参考输入关键点和座椅的预设兴趣区域(ROI)来将乘客匹配到座位,从而匹配车辆。并检测占用情况。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号