A method for detecting a wiping area of a vehicle wiper based on visual measurement

Abstract

The application discloses a vehicle wiper wiping area and frequency detection method based on a structured light camera. The method collects a windshield image sequence, a depth map and a collection time through the structured light camera; based on motion detection, each frame of motion area mask is extracted, and a cumulative wiping area mask is obtained through multi-frame fusion; three-dimensional point cloud is projected back through depth data, a windshield curved surface is fitted, and after region mapping, the curved surface area, position and shape are obtained through parameter coordinate conversion and grid accumulation. A time sequence is constructed by extracting an attitude angle based on a single frame mask, and a reference angle is determined by a first stroke; when the angle of a subsequent stroke crosses the reference angle, a reference time is obtained by using adjacent sampling point interpolation compensation, and a wiping frequency is calculated according to an integer complete round trip period. The method considers the windshield curvature, synchronously detects the wiping area and the frequency, and improves the accuracy, stability and engineering applicability.