Designing And Implementation Of The Image Acquisition And Processing System For Industrial Line Scan Cameras

With the rapid development of machine vision systems, the image data acquisition andpre-processing accuracy have become increasingly demanding, but now the image dataacquisition and pre-processing system design can not meet the industry and science needs.In particular, many of the current complex machine vision testing system’s testing accuracyis low, so the image pre-processing system based on the high-speed data acquisition isrequired. Also in the conventional machine vision and image analysis have been use of PCdone the image algorithm processing, but because of the image processing low speed, it isdifficult to meet the speed requirements of today.Based on this consideration, this paper intends to research high-speed imageacquisition measurement system based on high-resolution linear CCD industry camera oflarge three-dimensional measurement system, with an emphasis on high-speed imageacquisition and pre-processing technology.The main subject of the work is completed the following two points:1)In the First, in the basis of studying the Camera Link standard and focuses on itsthree operating modes, designing the front-end data acquisition interface circuit based onCamera Link standard, mainly through the interface daughter board’s (DS90CR288A) levelconversion function converts the low-voltage differential signal to the CMOS/TFL(LVCMOS/LVTTL) signals that FPGA can normally receive, also achieved the image dataacquisition card serial communication between the camera and greatly improving theaccuracy of the image data acquisition. The entire image data acquisition andpre-processing system use DE4development board as the main controller, and the imagereceiving sub-board use the ALTERA production’s CLR_HSMC board. After testing, thismodule complete the function of converting differential signal to CMOS signal, cameradata’s acquisition and image display and communication control.2) The second, in the studying of the pretreatment portion of the image, first we willintroduce the basic concepts of image edge detection and judgment of the Gaussiandistribution, and research and compare a variety of image edge detection algorithm,focusing on a variety of edge detection operator comparing to choose the method bestsuited for hardware implementation. Here we choose Sobel edge detection algorithm as themain image operator. And for the current via PC host computer a digital image edgedetection’s processing speed is slow, design an implementation plan based on Sobel Edge Detection Hardware Description achieved. Through simulation and hardware integrated,comparative results verify that the program can be completed in the basis of ensuring theaccuracy of edge detection, and further enhance the image edge detection rate. In the studyof the control of the rotation angle of the camera, the subject will be used to control PTZand coding controller output, based on this, we designed a PTZ coding controller designand simulation implementation.In the integrated overall system, we will use the Altera companies’ Avalon bus toconnect each subsystem, the operation of the overall system of coordinated control throughNios CPU as the host controller, the final output of the final results of the edge detectioncontroller to the PC via the serial UART port. Therefore, this paper also focuses on thebasic concepts and knowledge about Avalon agreement.