Image Deblurring Parallel Processing System Based On Dual DSP

With the development of the era of industrial informatization,the vision of machine is more and more widely applied in various fields,and image deblurring technology is an important branch of it.The application of image deblurring technology demand is growing in production and living world.Due to camera shake itself and focus inaccurate,or the effects of factors such as atmospheric disturbances,will cause the captured image distortion on digital camera,the specific performance is blurring.Unified image deblurring algorithm can deblur a variety of causes of blur image restorations into a clear image.Compared with other image processing algorithms image deblurring algorithm is complex,and the develop process is difficulty.Currently,the existing image deblurring technology tends to stay on the simulation of PC platform stage,it is using computers to realize the function of image restoration.Due to the high cost,large volume,and large power consumption on PC,these disadvantages of greatly limit the image deblurring technology application in real life.The best way to solve these problems is transplanted the image deblurring algorithm to the embedded chip which to make the function of algorithms to be miniaturization.This subject adopts TMS320C6657 high-performance dual-core DSP launched by TI company as the computing chip of image deblurring,firstly analyzes the hardware architecture of digital image processing platform,describes the framework of software engineering combined with the DSP chip structure.Then introduce the transplantation of image restoration algorithm and code debugging method in detail,and make precise statistics and analysis to the execution time in the algorithm after transplanted.The image restoration algorithm is optimized by a variety of methods and the optimization is turn out to be feasibility and efficiency by experiments and simulation methods.In combination with many kinds of communication mechanisms in SYS/BIOS operation system,the image deblurring parallel algorithm on the dual-core DSP are studied,and realized the dual parallel processing.Then the data processing flowing on the system hardware platform are analyzed in detail,and the system experiment realized the image data receiving,algorithm processing,sending of the results of the output function.Finally introduced the infrared gas detection system based on DaVinci DM8148 DSP platform,and obtained a batter result.Through this topic research basically realized the function of image restoration on DSP chip,the whole hardware system is become miniaturization,micromation,and the application of image restoration technolog expands widly,provided the reference for subsequent research.