Скачать или смотреть Demo Real-Time-Yolov8-Pose-tracking-Sequence (GitHub)

Human Pose tracking has always been a very important and interesting topic in the field of computer vision. Through pose tracking, we can understand the change of human or object's pose. YOLOv8-Pose provides great pose tracking implementation. Yolov8-Pose pose tracking technology parses 17 keypoints of data (x,y,visibility) of each person from continuous images (video/streaming) and gives an ID to each person, which is more valuable if the pose data and tracking results can be utilized efficiently.
Here we provide an algorithm to implement the human skeleton sequence.

姿態追蹤一直以來都是機器視覺領域中一個非常重要且有趣的議題。透過姿態追蹤可以了解人或物件的姿態變化。在YOLOv8-Pose提供非常棒的姿態追蹤實現。透過Yolov8-Pose姿態追蹤技術將從連續影像(影片/串流)當中解析出各個人員的17個關鍵點數據(x,y,visibility)以及賦予每個人員一個ID。如果能夠有效利用姿態數據以及追蹤結果將有更高的利用價值。
這裡提供一個演算法實現人體骨架序列。