This study proposes a new method for multiline laser classification and optical plane calibration to overcome the limitations of the existing on-line laser calibration methods used for three-dimensional (3D) measurement systems. These limitations include difficulties in multiline calibration and low accuracy in extracting feature points. The proposed method utilizes the steger algorithm to extract the laser center line and classifies it using a method based on contour judgment. The key points on the calibration board are detected, and the corresponding relationship matrix between the camera and calibration board is obtained using singular value decomposition. Using this relationship matrix, the 3D coordinates of the key points in the camera coordinate system can be determined. Subsequently, the plane equation of the calibration plate in the camera coordinate system is obtained via fitting, allowing for the determination of the 3D point of the laser line on the calibration plate plane. This process can be repeated several times by changing the position and attitude of the calibration plate till multiline laser plane calibration is achieved. Experimental results on the standard block demonstrate that the proposed method is effective and meets the industrial accuracy parameters, achieving an average size accuracy of 0.02 mm after repeated measurements.