An intelligent human-computer interaction method based on dynamic gesture recognition

Abstract

The invention discloses an intelligent human-computer interaction method based on dynamic gesture recognition, belonging to the technical field of human-computer interaction. The invention generally adopts the traditional machine learning algorithm for the current gesture-based human-computer interaction scheme, and cannot meet the requirements of real-time and robustness of the embedded equipmentunder the background of a smart home. By improving the lightweight target detection network, the invention realizes the fast and accurate detection of the hand region, on the basis of which the target tracking algorithm is integrated to obtain the motion trajectory of the hand and the personalized human-computer interaction behavior is provided according to the trajectory classification result. The invention can realize the real-time recognition of the dynamic gesture on the embedded device and has strong robustness to factors such as illumination, skin color, background and the like, and isan intelligent human-computer interaction solution facing the intelligent family scene.

Description

technical field [0001] The present invention proposes a non-contact interaction method based on dynamic gesture recognition, which is an intelligent human-computer interaction solution for smart home scenarios. Background technique [0002] In recent years, with the rapid development of machine learning technology, especially deep learning theory, and the increasing maturity of hardware equipment manufacturing processes such as high-definition cameras and high-performance graphics cards, artificial intelligence technology has been mainly used in the industry and has begun to be used more and more. It is widely used in people's daily life, for example, intelligent access control system based on license plate recognition, unattended supermarket based on face recognition, face cartooning software based on style transfer, etc. In this context, the "smart home" concept, which aims to integrate artificial intelligence, the Internet of Things, and cloud computing to create a comfor...

AI Technical Summary

Problems solved by technology

[0003] At present, there are already a batch of human-computer interaction solutions and equipment based on gesture recognition on the market. However, such solutions generally only recognize static gestures, and most require special equipment, and the application scenarios are relatively limited.

Take the gesture operation of Microsoft’s Xbox One console as an example. Its gesture interaction solution not only needs to be equipped with a dedicated Kinect depth camera, but basically can only recognize body movements. The recognition rate of gestures is low, which cannot meet the daily needs of users.

On the other hand, due to the limitation of the computing power of embedded devices in the home scene, existing solutions cannot use convolutional neural networks with the best performance in the industry, such as residual networks, but can only use support vector machines in traditional machine learning algorithms. Even simple template matching methods are used for static gesture recognition. On the one hand, such methods have poor recognition effects, especially poor robustness to factors such as illumination, skin color, gesture direction, and image background; on the other hand, such methods It is not easy to integrate hand area tracking and trajectory classification algorithms, and it is impossible to realize interactive response behaviors to dynamic gestures, that is, gesture movement trajectories, and basically cannot meet the needs of human-computer interaction in the context of intelligent homes

Generally speaking, facing the era background of home intelligence, traditional machine learning methods can no longer meet the requirements of robustness, real-time, and accuracy on embedded devices for intelligent human-computer interaction solutions. Therefore, the introduction of deep learning solution, implementing a deep convolutional neural network on embedded devices is an inevitable requirement for the development of smart homes

Images