Search

CN-115439552-B - Multi-camera external parameter calibration method and device and electronic equipment

CN115439552BCN 115439552 BCN115439552 BCN 115439552BCN-115439552-B

Abstract

The application discloses a multi-camera external parameter calibration method and device and electronic equipment, and relates to the technical field of intelligent driving. The multi-camera external parameter calibration method comprises the steps of determining initial external parameters corresponding to each camera respectively, performing perspective transformation on first images acquired by each camera according to perspective transformation matrixes corresponding to the initial external parameters to obtain second images, identifying all first target angular points of a calibration plate in the second images, and performing joint adjustment on the initial external parameters based on first image coordinates corresponding to the first target angular points respectively. By the method, the acquired images are subjected to perspective transformation operation, so that the acquired images are subjected to visual angle correction, the recognition accuracy of the corner points of the calibration plate is further improved, and the accuracy of external parameter calibration of each camera is higher.

Inventors

  • DU WEISONG
  • WEN WEIJIN
  • LIU DEXUN

Assignees

  • 浙江吉利控股集团有限公司
  • 宁波吉利汽车研究开发有限公司

Dates

Publication Date
20260512
Application Date
20220824

Claims (9)

  1. 1. A multi-camera extrinsic parameter calibration method, the method comprising: determining initial external parameters corresponding to each camera respectively; According to the perspective transformation matrix corresponding to each initial external parameter, performing perspective transformation on the first image acquired by each camera to obtain each second image; Identifying all first target corner points of the calibration plates in each second image; based on the first image coordinates corresponding to each first target angular point, carrying out joint adjustment on each initial external parameter, including: determining first image coordinates corresponding to each first target angular point respectively; According to the arrangement sequence among the first target angular points, determining second world coordinates corresponding to the first image coordinates respectively; inputting a preset model according to each second image coordinate and each second world coordinate to obtain optimization parameters corresponding to each camera respectively; And carrying out joint adjustment on each initial external parameter according to each optimization parameter.
  2. 2. The method of claim 1, wherein determining initial extrinsic parameters for each respective camera comprises: performing de-distortion operation on the third images acquired by each camera to obtain fourth images; determining second image coordinates corresponding to preset corner points of the calibration plates in each fourth image; according to the arrangement sequence among all preset angular points, determining first world coordinates corresponding to all second image coordinates respectively, wherein the first world coordinates are three-dimensional reference coordinates; And determining the initial external parameters corresponding to each camera respectively according to each second image coordinate and each first world coordinate.
  3. 3. The method of claim 1, wherein performing perspective transformation on the first image acquired by each camera according to the perspective transformation matrix corresponding to each initial external parameter to obtain each second image comprises: Calculating a perspective transformation matrix corresponding to each camera according to each initial external parameter; and performing perspective transformation on the first images acquired by each camera through the transformation matrix to obtain each second image.
  4. 4. The method of claim 1, wherein identifying all first target corner points of the calibration plate in each second image comprises: identifying all corner points of the calibration plate in each second image; Determining abnormal corner points corresponding to a common view area from all the corner points, wherein the common view area is an overlapping area of acquisition ranges between two cameras; and eliminating the abnormal corner points from all the corner points to obtain all the first target corner points.
  5. 5. The method of claim 4, wherein determining outlier corners corresponding to the common region among the all corners comprises: determining each first common view point corresponding to the common view area in each second image; matching each first common view point with each second common view point corresponding to the common view area in the world coordinate system; if the corner points which cannot be successfully matched exist in the first common view points, determining the corner points which cannot be successfully matched as abnormal corner points.
  6. 6. A multi-camera extrinsic calibration apparatus, the apparatus comprising: the determining module is used for determining initial external parameters corresponding to each camera respectively; The transformation module is used for performing perspective transformation on the first images acquired by each camera according to the perspective transformation matrix corresponding to each initial external parameter to obtain each second image; The identification module is used for identifying all first target corner points of the calibration plate in each second image; The adjusting module is configured to jointly adjust each initial external parameter based on first image coordinates corresponding to each first target corner, and includes: determining first image coordinates corresponding to each first target angular point respectively; According to the arrangement sequence among the first target angular points, determining second world coordinates corresponding to the first image coordinates respectively; inputting a preset model according to each second image coordinate and each second world coordinate to obtain optimization parameters corresponding to each camera respectively; And carrying out joint adjustment on each initial external parameter according to each optimization parameter.
  7. 7. The apparatus of claim 6, wherein the determining module is specifically configured to: performing de-distortion operation on the third images acquired by each camera to obtain fourth images; determining second image coordinates corresponding to preset corner points of the calibration plates in each fourth image; according to the arrangement sequence among all preset angular points, determining first world coordinates corresponding to all second image coordinates respectively, wherein the first world coordinates are three-dimensional reference coordinates; And determining the initial external parameters corresponding to each camera respectively according to each second image coordinate and each first world coordinate.
  8. 8. An electronic device, comprising: a memory for storing a computer program; a processor for carrying out the method steps of any one of claims 1-5 when executing a computer program stored on said memory.
  9. 9. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored therein a computer program which, when executed by a processor, implements the method steps of any of claims 1-5.

Description

Multi-camera external parameter calibration method and device and electronic equipment Technical Field The application relates to the technical field of intelligent driving, in particular to a multi-camera external parameter calibration method and device and electronic equipment. Background With the rapid development of intelligent driving automobiles, the automobile industry has a rapid increase in the demand for auxiliary driving functions under visual schemes, so that a 360-degree, 540-degree and other multi-camera looking-around system has become the conventional demand of intelligent automobiles, and parameter calibration on multiple cameras is the basis of the looking-around system, and the final effect of the auxiliary driving system is directly influenced by the quality of the parameter calibration. Generally, the camera parameter calibration process is divided into two parts, one part is converted from a camera coordinate system to an image coordinate system, and the other part is converted from a three-dimensional point to a two-dimensional point, including camera internal parameters, wherein the camera internal parameters are descriptions of physical characteristics of the camera, such as focal length, separation rate and the like. Another part is the conversion from world coordinates to camera coordinates, which is the conversion of three-dimensional points to three-dimensional points, including camera outliers, which determine the position and orientation of the camera in a certain three-dimensional space, such as rotation parameters, translation parameters, etc. In the process of camera external parameter calibration, the calibration plates are all tiled on the ground, a certain angle exists between the shooting visual angles of all cameras of the looking-around system and all the calibration plates, so that a stretching phenomenon can occur in the shot calibration plate image, for example, the edge line of the calibration plate close to the camera is relatively longer, and the edge line of the calibration plate far from the camera is relatively shorter. Under the condition, the acquired calibration plate image and the standard image have larger difference, so that the subsequent identification precision of each angular point in the calibration plate is seriously influenced, and the external parameter calibration of each camera is inaccurate. Disclosure of Invention The application discloses a multi-camera external parameter calibration method, a device and electronic equipment, which are characterized in that through performing perspective transformation operation on images acquired by each camera, the recognition accuracy of the corner points of the calibration plate is improved, and the accuracy of external parameter calibration of each camera is further improved. In a first aspect, the present application provides a method for calibrating a multi-camera external parameter, the method comprising: determining initial external parameters corresponding to each camera respectively; According to the perspective transformation matrix corresponding to each initial external parameter, performing perspective transformation on the first image acquired by each camera to obtain each second image; Identifying all first target corner points of the calibration plates in each second image; And based on the first image coordinates corresponding to each first target angular point, carrying out joint adjustment on each initial external parameter. By the method, the acquired images are subjected to perspective transformation operation, so that the acquired images are subjected to visual angle correction, the recognition accuracy of the corner points of the calibration plate is further improved, and the accuracy of external parameter calibration of each camera is higher. In one possible design, the determining initial parameters corresponding to each camera includes: performing de-distortion operation on the third images acquired by each camera to obtain fourth images; determining second image coordinates corresponding to preset corner points of the calibration plates in each fourth image; according to the arrangement sequence among all preset angular points, determining first world coordinates corresponding to all second image coordinates respectively, wherein the first world coordinates are three-dimensional reference coordinates; And determining the initial external parameters corresponding to each camera respectively according to each second image coordinate and each first world coordinate. By the method, initial external parameters corresponding to each camera are calculated preliminarily, and a basis is provided for external parameter adjustment of each subsequent camera. In one possible design, the performing perspective transformation on the first image acquired by each camera according to the perspective transformation matrix corresponding to each initial external parameter to obtain each second image