Image Processing Of Pavement Research Based On Embedded System

The embedded image detection technology is also playing more and more vital role in the pavement surface detection.This article unifies the embedded system development technology, image processing technology and the pavement detection technology.It aims at the embedded system. The embedded system is small, the power loss is low, the speed is quick.So the pavement embedded image processing system is designed.This paper summarizes the ARM processor embedded system characteristics and current developments, analysis the each module which is designed to achieve the principles.Select Mini2440 development board which is based on Samsung's S3C2440 processor as the pavement image processing system hardware platform. According to actual needs, analysis and compare a variety of embedded operating system. So decided to use Windows CE as a platform operating system. In the Mini2440 hardware platform, Windows CE operating system is reduced and transplanted.This article describes how to use Visual Studio 2005 and ActiveSync set up development environment. This system's graphical user interface and some of the main interface are made. The system realizes the gray image processing, image enhancement and image segmentation algorithms, each running results is also given in the graphical interface.In the aspects of the image data processing, this system use a special data structure to store the gray scale image data, so that it can reduce the amount of data processing. The system also use the image sharpening to complete a pavement crack image information processing.This system use class discriminant analysis method to accomplish the image threshold segmentation.The system completed the extraction of the road surface area, perimeter, length, average width, the percentage of cracks.This system can be applied to automatic detection of damaged road, the system can also provide the scientific data to the road management and maintenance department.