Automated Radar Mount-Angle Calibration in Automotive Applications

Abstract

Precise knowledge of alignment and orientation of radar sensors around the vehicle is imperative to determining location of objects that it sees in ADAS applications. Experience has shown that erroneous sensor alignment and orientation leads to biased object location estimates, which manifest itself in incorrect decision making. Determining the location and orientation of sensor with respect to an origin on the vehicle is referred to as sensor calibration. Though many have acknowledged the problem of sensor calibration and emphasized on the need for its solution, to the best of our knowledge, there is no method publicly available to address this issue. In this paper, we present a new method for sensor calibration that is easy to execute, is repeatable with minimal supervision, and yields close-to-accurate results. We have demonstrated the viability of proposed calibration method by using the results obtained in an occupancy grid mapping application.