Behavior recognition and action detection recognition method and system based on double-flow CNN

The embodiment of the invention provides a behavior recognition and action detection recognition method and system based on a double-flow CNN, and belongs to the technical field of computer vision. The identification method comprises the following steps: acquiring optical flow information and image...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: YANG NAIQI, CHEN YING, ZHU ZHONGXIAN, GUO LONGGANG, WEI NAN, ZHU YUANFU, GUO ZHENYU, WANG YUN, CAI KEWEI, WENG LING, MA HUAN, PU DAOJIE, DU YAO, CHANG WENJING, DU PENG, LIU XIN, WU XIANG, SHEN GUOTANG, SHI WEN, WANG WEIWEI, HUANG DAOJUN, XU MENGFU, ZANG CHUNHUA, LIU WENTAO, SHAO HUA, ZHANG XUEYOU
Format: Patent
Sprache:chi ; eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The embodiment of the invention provides a behavior recognition and action detection recognition method and system based on a double-flow CNN, and belongs to the technical field of computer vision. The identification method comprises the following steps: acquiring optical flow information and image information of field camera equipment; according to the optical flow information and the image information, a convolutional neural network is adopted for calculation, so that the recognition probability of the field action is obtained; and selecting the field action with the maximum identification probability average value as an identification result. According to the technical scheme, the behavior recognition and action detection recognition method and system based on the double-flow CNN, provided by the invention, perform calculation and distribution by collecting the optical flow information and the image information of the site and combining the two neural networks, and realize accurate detection of behaviors a