Image Recognition And Application Of Linear Object

The road extraction in image is a very important research interests in field of digital image processing, and there are good prospects. The road information has a very widely use in many fields. However, due to the complexity of the image scene, road features are subject to many disturbances, the road extraction of image has not yet reached the requirement of practical application.The research of road Extraction have many years of history, have made many valuable algorithms, however, these algorithms have some limitations. This paper summarizes and analyzes some typical algorithms in the field of road extraction that the previous proposed, and then, designed a series of algorithms to extract road. In this paper, we extract two kind of road image: simple road image and high-resolution images of urban roads. The main contents of the paper are as follows:(1) We have done some in-depth research on the model of road extraction, analysis the characteristics of road image, present various factors that impact the road extraction result. Establish a system for extracting road in scanned map image based on the analysis of the feature in the scanned map image. The system contains crossing identification, noise removal, road refinement, binary, coordinate store. Experimental results show that the system has some practical value.(2) After studying the high-resolution of urban roads, we design a series of algorithms to extract the road in the image, the algorithm take the characteristics of the road image into account. In the extraction process, we use watershed dual-thresholds algorithm, multi-weighted method, mathematical morphology, shape index step by step.