An Automatic Evaluation System for the Photometric Performance of Vehicle Headlamps Using Image Processing Algorithms

Global report reveals that about 1.24 million road traffic deaths occur annually on the world's roads. The beam pattern is an important characteristic of the headlamp for road safety. Measurement by traditional testing methods using a CIE A-alpha type goniophotometer cost enormous amount of time and money. In this study, an automatic evaluation method based on image processing is proposed for saving testing time and space. The novel concept of I-matrix is applied in the evaluation algorithm. Comparison of six different algorithms for edge detection of the cut-off line is shown in this study and the Canny algorithm stands out with an expected result. Experiment results show that the proposed method can export a reliable report with a tolerant error. The automatic alignment procedure based on edge detection algorithms can not only serve this evaluation system, but also be used to replace the visual task for cut-off line aiming in traditional methods.