A gesture recognition method and system based on a neural network

Abstract

The invention provides a gesture recognition method and system based on a neural network, and the method comprises the steps of building a training sample set, and carrying out the size normalizationprocessing of an image of a training sample; classifying the gestures in the normalized image, and making different segmentation tags; Enhancing the classified gesture images, constructing an FCN network model, and training the FCN network model by using the enhanced images; according to the method, optimizing the classification quality and the segmentation quality of the trained FCN network model, using the optimized FCN network model for identifying the collected samples, so that the method and the system are excellent in performance on a test data set, and the classification effect is better even if the method and the system are directly applied to other gesture libraries or living scenes without training.

Description

technical field [0001] The present disclosure relates to a gesture recognition method and system based on a neural network. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] With the in-depth application of computer technology, the demand for human-computer interaction (HCI) technology is also increasing. However, traditional mechanical interaction methods such as keyboard and mouse are inconvenient in many scenarios due to the need for touch operation and adaptation process. With the continuous humanization and intelligence of computers, and the continuous development of user experience and scene interaction requirements, some human-computer interaction technologies that conform to human habits, such as speech recognition, face recognition, eye tracking, human body posture recognition, and gesture recognition, are becoming more and more p...

AI Technical Summary

Problems solved by technology

Its advantages and disadvantages are: the classification features and classifiers are artificially designed, highly targeted, and can fully combine the shape of the gesture itself

Its advantages and disadvantages are: the class features and classifiers are obtained by the network self-learning, and the environment adaptability is good, but the prior knowledge of gesture shapes cannot be effectively used

Therefore, it has strong adaptability to application scenarios, but is susceptible to interference from complex backgrounds

[0008] In addition, gesture recognition methods that combine the two methods have also appeared in the past two years, but most of them have failed to effectively overcome the technical problems in the above two types of methods and need to be improved.

Images