Precision-adjustable neural network camera calibrating method

Abstract

The invention is a kind of calibration method for nerve network camera with adjustable accuracy. It divides the image according to the radial deformation degree, gets a series of concentric circles as the training area of the nerve network from the initial radius, changes the amounts of division according to the deformation degree, it divides the image fine in the serious deformation area far away from the centre, and it divides the image roughly in the mild deformation area near the centre, in order to decrease the error caused by radial deformation; divides the calibrated image, uses the international coordinate as the input, the image is used as the output, forms the three outputs, two inputs BP nerve network, and trains the divided area.

Description

Technical field: [0001] The invention relates to a neural network camera calibration method with adjustable precision, which is used for calibrating the camera. It belongs to the field of advanced manufacturing and automation technology. Background technique: [0002] The basic task of computer vision is to calculate the three-dimensional geometric information of the object from the image information obtained by the camera. The position of each point on the image is related to the geometric position of the corresponding point on the surface of the object. The interrelationship of these positions in the image is determined by the camera imaging geometry. The parameters of the model are called camera parameters, including determining the internal geometric and optical parameters of the camera (internal parameters) and the relationship between the camera and the world coordinate system (external parameters). These parameters must be determined experimentally, a process known ...

AI Technical Summary

Problems solved by technology

Some literatures have taken this factor into consideration, but they have not clearly divided the training area of ​​the neural network. They simply divide the image area artificially into the central area and non-central area of ​​the image based on experience, and consider the parts outside the central area of ​​the image to be The degree of distortion is the same. This method improves the recognition rate of the system to a certain extent, but it still cannot fundamentally solve the problem that the farther away from the center of the image, the greater the severity of the distortion. Simply divide the outer layer of the image As a class, there is still a large error

Images