Effective Photometric Alignment for Surround View Monitoring System

Kwon Lee, Deukhyeon Kim, Hoseok Do, Jihyun Kim, Kyuyeol Chae

Abstract: Surround view monitoring (SVM) system provides a composite bird-eye view of the vehicle to assist in safe parking. Since each camera independently performs auto exposure (AE) and auto white balance (AWB), the composite bird-eye view has noticeable boundaries between adjacent views. To achieve a seamlessly stitched bird-eye view, we propose an effective photometric alignment for surround view monitoring system using a simple additive gain model. Experimental results show that the proposed method provides a seamlessly stitched bird-eye view. And the processing time of the proposed method achieves 3ms on the NVIDIA Tegra CX embedded platform.