Automatic Calibration of an Around View Monitor System Exploiting Lane Markings

Abstract

This paper proposes a method that automatically calibrates four cameras of an around view monitor (AVM) system in a natural driving situation. The proposed method estimates orientation angles of four cameras composing the AVM system, and assumes that their locations and intrinsic parameters are known in advance. This method utilizes lane markings because they exist in almost all on-road situations and appear across images of adjacent cameras. It starts by detecting lane markings from images captured by four cameras of the AVM system in a cost-effective manner. False lane markings are rejected by analyzing the statistical properties of the detected lane markings. Once the correct lane markings are sufficiently gathered, this method first calibrates the front and rear cameras, and then calibrates the left and right cameras with the help of the calibration results of the front and rear cameras. This two-step approach is essential because side cameras cannot be fully calibrated by themselves, due to insufficient lane marking information. After this initial calibration, this method collects corresponding lane markings appearing across images of adjacent cameras and simultaneously refines the initial calibration results of four cameras to obtain seamless AVM images. In the case of a long image sequence, this method conducts the camera calibration multiple times, and then selects the medoid as the final result to reduce computational resources and dependency on a specific place. In the experiment, the proposed method was quantitatively and qualitatively evaluated in various real driving situations and showed promising results.

An Automatic Calibration Method for AVM Cameras

“…The Hough transform processes all the linear components that pass through a point into distance and angular components from the midpoint , and this gives us an advantage in that a straight line through which the most points pass can be detected. Equation 18 Hough transform, where x and y denote image coordinates, θ the slope of the straight line, and r the distance between the origin and the straight line [13,24]. One has…”

“…Because the parking line has a thick line component, it has two components: a rising edge, which represents the intensity change from dark to bright, and a falling edge, which rep- resents the intensity change from bright to dark. The AVM image can be corrected again using the rising and falling edges [13]. Figure 16 depicts the Hough-transformation result.…”

“…The camera-movement angle can be measured using the location of the vanishing point. Equations (19) and (20) show the relationship between the vanishing point and cameramovement angle [13]. One has…”

“…In this study, we focus on an AVM system, which comprises cameras in the front, rear, left, and right of the vehicle, respectively, to display the 360°view of the surroundings of the vehicle. Additionally, we focus on constructing a system that assists in parking by synthesizing images from the cameras to from a topview image [1][2][8][9][10][11][12][13]. However, AVM cameras must be calibrated to synthesize top-view images.…”

Special Issue on Visual Sensors

“…The proposed approach is 1.5 times more precise than using standard calibration with a checkerboard pattern. Finally, Choi et al [19] proposed a method that automatically calibrates four cameras of an around view monitor system in a natural driving situation. Object recognition is a task in which a vision system is almost always involved.…”