This paper deals with a vehicle detection method for realizing a blind spot warning function under various environmental conditions. We introduced a method that is capable of discriminating the target detection objects under dark to bright conditions, and furthermore under various road conditions and occlusions that the image sensor may be affected. The system consists of five functional components: obstacle detection, velocity estimation, vertical edge detection, and final classification. In obstacle detection, a predicted image is generated under the assumption that the road surface is a perfectly flat plane, and then an object is detected based on a histogram that is created by comparing the predicted image and an actually observed image. The paper describes methods to enhance the performances under night conditions and also suppress false detections caused by residue on the lens, which becomes essential for installation on actual road conditions.