Adaptive calibration technique for dynamic character based on surveillance video

作者：Qiao Tiezhu; Chen Xin^*; Wang Feng; Jin Baoquan

来源：Chinese Journal of Scientific Instrument, 2014, 35(5): 1086-1092.

• Adaptive calibration
• Calibration algorithm
• Distortion parameters
• Dynamic monitoring
• Non-linear optimization
• Satellite positioning
• Surveillance video
• Three dimensional coordinate

In order to improve the adaptability and flexibility of camera calibration, an adaptive calibration method for dynamic character based on video surveillance using the basic principle of pinhole optical imaging. The method combines Faugeras calibration algorithm and satellite positioning principle to perform corner detection and contour point extraction for the character in binocular video after background elimination. Then, the corresponding corner points in the binocular image and adjacent frame images are matched and substituted into the three-dimensional coordinate adaptive calibration formula to form the calculation ring area of the camera parameter calibration and obtain the inside and outside parameters of the visual sensor; the method of linear least squares is used to solve the distortion parameters; global nonlinear optimization is performed on the calculated results and the optimized internal and external parameters of the camera are obtained. Finally, according to the pixel information of the characteristic points of the dynamic character, Zhang Zhengyou calibration algorithm and the proposed method were compared, their relative RMS measurement errors of the overall parameters were analyzed, and the height of the dynamic character was solved. This method does not require knowing any camera parameters, and is simple and convenient. Experiment results show that the method has high calibration accuracy and robustness.

• 出版日期2014
• 单位太原理工大学