Camera calibration is critical in computer vision measurement system, affecting the accuracy of the whole system. Many camera calibration methods have been proposed, but they cannot consider precision and operation co...Camera calibration is critical in computer vision measurement system, affecting the accuracy of the whole system. Many camera calibration methods have been proposed, but they cannot consider precision and operation complexity at the same time. In this paper, a new technique is proposed to calibrate camera. Firstly, the global calibration method is described in de-tail. It requires the camera to observe a checkerboard pattern shown at a few different orientations. The checkerboard corners are obtained by Harris algorithm. With direct linear transformation and non-linear optimal algorithm, the global calibration pa-rameters are obtained. Then, a sub-regional method is proposed. Those corners are divided into two groups, middle corners and edge corners, which are used to calibrate the corresponding area to get two sets of calibration parameters. Finally, some experimental images are used to test the proposed method. Experimental results demonstrate that the average projection error of sub-region method is decreased at least 16% compared with the global calibration method. The proposed technique is simple and accurate. It is suitable for the industrial computer vision measurement.展开更多
基金Tianjin Research Program of Application Foundation and Advanced Technology(No.14JCYBJC18600,No.14JCZDJC39700)the National Key Scientific Instrument and Equipment Development Project(No.2013YQ17053903)
文摘Camera calibration is critical in computer vision measurement system, affecting the accuracy of the whole system. Many camera calibration methods have been proposed, but they cannot consider precision and operation complexity at the same time. In this paper, a new technique is proposed to calibrate camera. Firstly, the global calibration method is described in de-tail. It requires the camera to observe a checkerboard pattern shown at a few different orientations. The checkerboard corners are obtained by Harris algorithm. With direct linear transformation and non-linear optimal algorithm, the global calibration pa-rameters are obtained. Then, a sub-regional method is proposed. Those corners are divided into two groups, middle corners and edge corners, which are used to calibrate the corresponding area to get two sets of calibration parameters. Finally, some experimental images are used to test the proposed method. Experimental results demonstrate that the average projection error of sub-region method is decreased at least 16% compared with the global calibration method. The proposed technique is simple and accurate. It is suitable for the industrial computer vision measurement.
