Body language recognition system based on neural network

Abstract

The invention discloses a body language recognition system based on a neural network, and the system comprises an obtaining module which is configured to obtain an image of a to-be-recognized object; the feature extraction module is configured to extract limb feature key points and hand feature key points from the image of the to-be-recognized object; deleting invalid images and key points corresponding to the invalid images according to set body language requirements in the key point extraction process; the data preprocessing module is configured to perform zero setting on the abnormal feature key points; the feature fusion module is configured to perform feature fusion on the limb feature key points and the hand feature key points to obtain fusion features; and the classification module is configured to classify the fusion features by adopting the trained classifier to obtain a body language recognition result. According to the method, invalid images are deleted and abnormal feature key points are subjected to zero setting operation, so that no noise of data is ensured, and a foundation is laid for accurate recognition of body languages.

Description

technical field [0001] The invention relates to the technical field of body language recognition, in particular to a body language recognition system based on a neural network. Background technique [0002] The statements in this section merely provide background related to the present disclosure and do not necessarily constitute prior art. [0003] Autism Spectrum Disorder (ASD) is a group of severe childhood neurodevelopmental disorders in which patients have difficulty developing normal social relationships and exhibit restricted or repetitive behaviors. Studies have shown that early detection and early intervention can significantly improve the prognosis of children with autism. Effective and accurate assessment methods can provide a reliable scientific and objective basis for the auxiliary diagnosis of autism and the later rehabilitation of children with autism. Clinically, doctors can interact with children through social-related games or actions, and then evaluate w...

AI Technical Summary

Problems solved by technology

[0004] Body language recognition in the prior art is to recognize body language for several specific environments or several specific application scenarios, for example, the camera intelligently recognizes whether people’s actions are dangerous behaviors; intelligently recognizes whether students raise their hands to speak in class, or whisper to each other in class Classroom-related behaviors such as whether you are serious about listening to lectures in class; intelligent recognition of people's emotional types, etc., but for the application scenario of intelligent recognition of social behavior actions, there is currently no accurate and efficient method that can be used

[0005] The existing technology can wear relevant sensors on children's limbs and then analyze and process the data collected by the sensors, and then perform body language recognition, but the recognition process has certain constraints on children, and equipment such as instruments also has a certain cost; moreover, The data collected by the existing technology has a lot of noise, and the accuracy of children's body language recognition is very low

Images