Optimizing the binocular vision calibration of the BP neural network based on a bat algorithm

• Main Authors: ZHAO YUHANG, QIAO YUJING, ZHANG SIYUAN
• Format: Patent
• Language: Chinese; English
• Published: 14.06.2019
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a visual measurement double-target calibration method, and belongs to the field of optical measurement and visual detection. The method comprises the following steps of determining the initial parameters such as the number of hidden layers of the BP neural network, input and output nodes and the like; determining a hidden layer node number range by utilizing an empirical formula, and determining an optimal hidden layer node number through a single-factor variance analysis method; according to a bat algorithm echo positioning principle, carrying out optimal selection onthe weight and the bias of the BP neural network, and determining an optimal weight and a bias value; and determining each parameter value of the BP neural network structure, enabling pixel point datadistribution to learn object space point data distribution, and completing calibration. According to the method, the number of hidden layer nodes is determined through a single-factor variance analysis method, and an error ge