Abstract: Aiming at the problems of local minimum and poor calibration precision of traditional BP(back propagation) network in the calibration process of binocular camera, an improved dung beetle optimizer (IDBO) was proposed to optimize the BP network (IDBO-BP) for binocular camera calibration.Firstly, the binocular camera calibration based on BP network was introduced. Then the dung beetle population was initialized by Piecewise chaotic mapping, which could increase the diversity and randomness of the population. Then, the rolling behavior of dung beetle was perturbed by the random walking strategy to increase the global optimization ability. Finally, the criss-cross strategy was used to perturb the whole population to improve the accuracy of the algorithm. The experimental results show that the calibration accuracy of BP network and IDBO-BP network for binocular camera is 0.1103 mm and 0.0256 mm, respectively,the calibration accuracy is improved by 76.8% . Therefore, the binocular camera calibration accuracy could be improved by the improved dung beetle optimization algorithm optimized BP network.