A road extraction algorithm for reducing shadow effect
Si Weian · 2021 IEEE Asia-Pacific Conference on Image Processing, Electronics and Computers (IPEC) · 2021
The traditional road shadow processing method distinguishes the road from the road shadow, which leads to the reduction of road automation. In this paper, based on the enhancement of the blue component, the brightness component in HSI space is transformed by NDUI shadow exponent transformation, the transformed image is segmented by threshold value, and the road main body is extracted after removing the small area spot and the adaptive structural element morphology closing operation. The test results were evaluated by the detection accuracy and quality, and it was concluded that the detection accuracy of the proposed method in rural roads could reach 96.87% and the detection quality could reach 85.21%. The test results show that this method can effectively eliminate the non-pavement shadow and extract the road with shadow, and has the characteristics of simplicity and high degree of automation.