An improved algorithm based on deep learning network for road image redundancy removal

Road detection is defined as one of the core technology of Advanced Driving Assistance System (ADAS), and this problem is important for improving the recognition accuracy and speed. Though much work has been done concerning road detection, the related questions about non-road areas are not thoroughly considered. Understanding the question is of primary importance in ADAS, we proposed an improved algorithm based on deep learning network for road image redundancy removal. Compared with the most typical road recognition methods, the experimental results show that the proposed method improves the speed of road recognition greatly, while ensuring the accuracy of road recognition to high level.

Automated summary

Road detection is crucial for ADAS, and our proposed improved algorithm utilizing deep learning network effectively improves the speed and accuracy of road recognition.