Research On Key Techniques Of Hardware System With Multi-camera

Image processing and video analysis technologies have been in existence for a long time.At the same time,the manufacturing processes of integrated circuits such as image sensors have also been continuously upgraded,and technologies and processes have complemented each other.Cameras have emerged as electronic products that can be seen everywhere in today's society.From the early film sensitization to the current use of image sensors,from low resolution to ultra-high resolution and from manual washing to automatic imaging,the development of the camera has been greatly promoted with the improvement of people's living standards.In recent years,the rapid development of technologies in Internet of things and mobile communication,the construction of smart cities and the variety of terminal photography devices in the consumer electronics market are changing greatly with each passing day.These all stimulate the upgrading of traditional cameras.At the same time,relevant researchers continued to specialize in research and achieved remarkable research results in related fields,making traditional cameras have significant improvements in image quality,field of view,and hardware appearance,including wide viewing.Field camera is one of its products.In the field of road traffic,all-round monitoring of roads without dead ends has always been the goal pursued by the people;the urban security field is also inseparable from the full range of video surveillance;in terms of consumer entertainment,with the rise of aerial photography of UAVs,taking the panoramic views of the region has become a big hobby for people.All of these highlight the wide-field camera application perspective.However,the multi-camera system is a mainstream scheme for wide field of view panoramic imaging technology.This article will design a multi-camera hardware processing system based on FPGA,CMOS and other chips.Meanwhile,the problem of jitter in video under harsh environment is studied and realized.The main work of this article is as follows:First,the hardware design of the multi-camera system is performed for the index requirements.Mainly starting from the overall system architecture,a series of key design points such as chip selection,chip-to-chip interconnection,lens selection,relative position deployment of the front-end camera and synchronization between cameras are elaborated in detail.From a theoretical point of view,the design of the hardware processing system is proved feasibility.Second,for the front-end camera in the multi-camera hardware processing system,RTL-level design and simulation were performed and finally realized.The front-end camera is the core of the whole wide-field multi-camera hardware processing system.The implementation flow of each module is introduced in detail and some simulation results are given.And the board-level verification passed,the final production of the board has been made of finished products.Thirdly,the phenomenon of imaging device blurring caused by rough roads or other harsh environments will be investigated.A video stabilization algorithm based on SURF feature points is studied and implemented at the RTL level.A block diagram and the simulation waveform diagram of key modules are given.After comparing and verifying the input video source with the processed output video,the effectiveness of the algorithm is proved.