Research And Design Of Video Image Acquisition System

With the development of information technology and the popularity of electronic equipment,the image acquisition device applies more and more extensively and the requirements of image obtained are getting higher and higher.In order to acquire the video sources for special requirements,it is often necessary to custom specific video image acquisition system.This thesis introduces the whole structure of the video image acquisition system.To get the video sources for image mosaic,our system provides two channels of video output and supports the real-time position adjustments in this design.Then the workflow and dataflow of video image acquisition system are analyzed detailedly from the features of analog input video sources and digital output sequence,and the design of hardware unites and software prototypes of our system are presented respectively.The hardware architecture of video image acquisition system is designed in two parts: video decoder unit and servocontrolled unit.SAA7111 and AL251 chips are used to convert the interlaced analog video signal into the progressive digital video signal,then the schematic and PCB layout of this video decoder board are drew by means of Altium designer software.Camera calibration module and stepper motor electronic control module are designed to achieve the multiple-degree-of-freedom and real-time adjustment of dual camera locations.Meanwhile,software video excitation generator is designed to verify the accuracy of the hardware output data and assertions are used to check the sequence in this design.In order to improve the image quality,a random-value impulse noise removal method based on switching weighted median filter is proposed after studying the noise features and image filtering methods.The algorithm has better noise detection capability,which can not only effectively remove impulse noise but also preserve image edges and details.And a verification platform based on UVM(Universal Verification Methodology)has been set up by developing our own components to verify the correctness of communication module in this system.The scoreboard component is utilized to automatically checks verification environment and the top random transaction-level cases is used to get the function coverage reports.Finally,the code coverage rates and function coverage rates reach 100% after test cases are adjusted several times.