Bulletin of Surveying and Mapping ›› 2023, Vol. 0 ›› Issue (4): 59-63,98.doi: 10.13474/j.cnki.11-2246.2023.0105

Previous Articles     Next Articles

Research and application of damage detection of highway billboard damage based on vehicle-mounted remote sensing image

ZHU Jianwei1,2, LI Chaokui1,2, ZHOU Xinshao3, ZHAO Dingying4, FU Kaihong1,2   

  1. 1. National and Local Joint Engineering Laboratory of Geospatial Information Technology, Hunan University of Science and Technology, Xiangtan 411201, China;
    2. Hunan Key Laboratory of Surveying, Mapping and Remote Sensing Information Engineering, Hunan University of Science and Technology, Xiangtan 411201, China;
    3. School of information and electronic engineering, Hunan City University, Yiyang 413000, China;
    4. School of Civil Engineering and Mechanics, Xiangtan University, Xiangtan 411201, China
  • Received:2022-06-24 Published:2023-04-25

Abstract: There is an inefficient and dangerous problem in the current manual inspection method of highway billboard management, and a new method of highway billboard inspection is proposed with the identification of damaged images of vehicle remote sensing billboards as the technical support. Based on the improved SIFT algorithm and the RANSAC algorithm, the method passes the discrete cosine transform before the extremum points of the SIFT Gaussian differential space are selected. Transform converts the light intensity data of the image into spectral data and performs a gating operation to screen out the high frequency coefficients while improving the quality of the initial set of inner points. Using RANSAC to calculate the perspective transformation model parameters for fine matching, and then adapt the matrix alignment to the opposite, the sum of the two images that are different from each other is determined, and the different connectivity domains are marked and extracted. Experimental results show that the performance of the proposed improved algorithm in the detection of billboard damage is robust, and the detection accuracy is more than 80% under the interference of various conditions, which meets the application requirements of this direction. The research results provide a new original technical support for Chinese highway management and related departments to identify, update and maintain highway billboards in a timely and effective manner.

Key words: vehicle remote sensing, SIFT, RANSAC, highway billboard, breakage detection

CLC Number: