This paper presents an improved feature detection and descriptor construction method for nighttime vehicle surrounding images. The proposed algorithm is based on the Oriented FAST and Rotated Brief (ORB) algorithms. In order to improve the accuracy of the proposed algorithm, the Contrast Limited Adaptive Histogram Equalization (CLAHE), feature detection, and descriptor construction were added. In addition, the proposed feature detection description method is realized by the Very Large-Scale Integration (VLSI) technique to meet the real-time application for vehicles. Compared with previous studies, the proposed design has better performance than previous methods of detecting features in nighttime for image stitching and recognition.