Human body upper limb intention recognition method based on projection reconstruction

Abstract

The invention discloses a human body upper limb intention recognition method based on projection reconstruction. The human body upper limb intention recognition method comprises the following steps: S1, acquiring a complete action track sample of human body upper limb movement, and constructing an original data set; S2, constructing an action initial stage data set; S3, performing projection and intention feature image reconstruction on the motion track sample in a three-dimensional space; S4, carrying out transfer learning and parameter adjustment on the Alexnet convolutional neural network; S5, inputting the motion track sample subjected to projection and intention feature image reconstruction in the step S3 into an Alexnet network subjected to transfer learning and parameter adjustment for training to obtain a human body upper limb motion intention recognition model; and S6, according to the human body upper limb motion intention identification model, identifying a human body upper limb intention in real human body upper limb motion. According to the method, a motion feature image is generated through projection and reconstruction operation of a human body upper limb motion track, and the Alexnet convolutional neural network is utilized to classify the motion feature image to realize accurate recognition of a human body upper limb motion intention.

Description

technical field [0001] The invention relates to the field of intention recognition, in particular to a method for recognizing intentions of human upper limbs based on projection reconstruction. Background technique [0002] In the wave of rapid industrial development, automated production is gradually shifting to intelligent and customized production. As an effective method to improve the intelligence and flexibility of automated production, the human-machine collaboration system provides a new solution for the development of industrial intelligence. In the human-machine collaboration system, using the motion data in the initial stage of the action to complete the recognition of the movement intention of the upper limbs of the human body can not only provide sufficient time for the robot to judge and plan, but also improve the work efficiency of the entire system. Recognizing human intent quickly and accurately is the basis for safe and effective human-robot collaboration. ...

AI Technical Summary

Problems solved by technology

These methods require the neural network to process each frame in the image sequence, which has the disadvantages of long calculation time and low accuracy.

Images