Depth Estimation Using Single Fisheye Camera

Abstract

<div>With the increasing utilization of automotive navigation systems in automotive platforms. The robot system is required to measure the distance of the surrounding area in real time. Omnidirectional distance measurement technology has emerged as a crucial requirement in the robotic industry. Traditional distance measurement methods for example stereo cameras and time of flight sensors have a limited field of view. This paper reports on a novel depth estimation method that operates on the mobile platform. The method involves collecting 180° fisheye images or 360° panoramic images via a single fisheye camera, detecting target objects in the image, and estimating their distance and size. This method can provide real-time omnidirectional distance estimation using a single fisheye camera, with an average recognition accuracy of 92.60% for 480 human targets with varying sizes and poses. The experimental result shows that the accuracy does not affect by the target size and pose. The improved distance estimation method has significant potential for robotic applications such as real-time obstacle avoidance.</div>