The Gesture Recognition Remote Controller Based On FPGA

Gesture recognition as an important tool for human computer interaction has been developed for more than twenty years.And it is also a hot topic for the scholars at present.Hand gesture recognition to control the household appliances not only can greatly simplify the control of household appliances, but also can provide users with intelligent operation experience. If the use of a gesture recognition function of the remote control replaces the ordinary remote control of home appliances, it will be a reform in the field of home appliances control. Therefore, gesture recognition remote controller has not only the theoretical research value, but also a broad market prospect. This paper makes use of the FPGA as the core processing unit of the gesture recognition system, which not only reduces the cost of the system, but also has good stability, and can satisfy the real-time processing of gesture recognition system.The gesture recognition system in this paper is based on Cyclone IV Altera series FPGA chip as the core processing unit. The system applies OV7725 camera modules to capture the image data, SDRAM chips to storage the image data, the VGA display to observe the process of the image processing and recognition results. Finally,the results of the gesture recognition will be transfered to the learning red remote control module and then control the household appliances.In the area of gesture recognition,11 kinds of static gestures and 4 kinds of dynamic gestures was defined in this paper to control the different functions of household appliances. First, the gesture recognition system will be acquired by the gesture image denoising filtering and color space conversion.Tthen through the color threshold segmentation and the rapid face detection method to remove the part of the face to get the segmentation. This method can effectively solve the problem of gesture segmentation in complex background. Next to the segmentation of the gesture with median filtering, according to the palm of the hand to intercept the key part of the palm.In addition, the system can detect what the type of the current gesture,a static gesture or a a dynamic gesture based on the position of the front and rear frame gestures. If it is a static gesture, it can be used as a feature extraction of gesture recognition. Then,matching the feature parameters of the standard gesture template library, we recognize the corresponding static gestures. If it is a dynamic gesture, the system will identify the corresponding dynamic gestures based on the motion trajectory of the front and rear frames. Finally, learning infrared remote control module could be driven by the system with the final identification results to control different household electrical appliances.The gesture recognition based on the design of this paper can be used without changing the structure of the original home appliance. The system can control many types of home appliances,even different manufacturers,which with the infrared control function,and has good real-time performance and accuracy.