abstract |
The invention discloses a human body posture detection method, device, equipment and storage medium. The method includes: collecting multiple frames of image data; inputting the current frame image data into a pre-trained human body posture detection model, so as to refer to the human body posture confidence map of the previous frame of image data, outputting a plurality of human body posture reference maps, and detecting the human body posture The model is generated by the training of the convolutional neural network applied to the embedded platform; the key points of the human body posture are identified in the human body posture reference map; the confidence map of the human body posture is generated according to the reliability of the key points of the human body posture; it is judged whether the current frame image data is is the last frame of image data; if not, input the human posture confidence map into the human posture detection model to participate in generating the human posture confidence map of the next frame of image data; if so, end the execution of generating multiple frames of image data. Manipulation of human pose confidence maps. The embodiment of the present invention realizes the human body posture detection on the embedded platform. |