作者: Roy McCann , Zhiqiong Hou , Weiguo Gong , Ron Strickland , Kelvin C P Wang
DOI: 10.3141/1968-01
关键词: Tracking system 、 Feature extraction 、 Traffic sign recognition 、 Kalman filter 、 Artificial intelligence 、 Dimensioning 、 Computer vision 、 Automation 、 Sign (mathematics) 、 Engineering 、 Image processing
摘要: Roadway signs represent a substantial investment of public money in road and highway infrastructure. However, the current level automation sign identification recognition, size dimensioning, location is unsatisfactory. In an effort to improve inventory, feature extraction Kalman filter-based tracking techniques for right-of-way (ROW) images are developed. A framework that combines conventional image-processing methods with filter method applied accuracy efficiency ROW image processing. With this technique, candidate region can be predicted on basis previous frame. processing used near sign, detection improved. The methodologies described fit dynamic moving environment, appropriate survey vehicle.