原文传递 IMAGE-PROCESSING METHODS FOR AUTOMATED DISTRESS EVALUATION.
题名: IMAGE-PROCESSING METHODS FOR AUTOMATED DISTRESS EVALUATION.
作者: Kalikiri-VK; Garrick-NW; Achenie-LEK
关键词: PAVEMENT-DISTRESS; DATA-COLLECTION; DATA-ANALYSIS; AUTOMATION-; IMAGE-PROCESSING; IMAGE-DATA-DIGITIZATION; BINARY-IMAGES; PATTERN-RECOGNITION
摘要: The collection and analysis of pavement distress data are the primary components of any pavement management system. There is an emerging trend over the last decade to automate this process of data collection and analysis. Current research work at the University of Connecticut aims at developing such an image data analysis system, compatible with the photolog system used by the Connecticut Department of Transportation for pavement distress surveys. The image data retrieved from the Connecticut photolog system cannot be used directly as input to the analysis system because of undesirable details such as shadows and random noise in the images. Therefore extensive image processing must be carried out before the image data are suitable for use as input for analysis. An image-processing strategy that can be used with the images obtained from the Connecticut photolog system is presented. Topics discussed include image data digitization, format conversion, feature extraction techniques, and binary image manipulation. The binary images resulting from the process were found to be a very good representation of the actual distress conditions of the pavement and can readily serve as input to a pattern recognition system for analysis.
总页数: Transportation Research Record. 1994. (1435) pp45-51 (2 Phot., 13 Fig., 10 Ref.)
报告类型: 科技报告
检索历史
应用推荐