Image-based weiqi manual identification method

Abstract

The invention discloses an image-based weiqi manual identification method. The method comprises a step of obtaining a weiqi board image and then correcting board information in the image, a step of separating whites and black chess pieces through an image processing method, a step of calculating a coordinate position of each placement point in a board, and a step of automatically extracting and judging a threshold through the calculation of a number of effective pixels of the vicinity of each placement point and a current condition, obtaining information whether the placement point has a whitepiece or a black piece, and obtaining manual data by recording the state information of all placement points. According to the method, the automatic identification of the board can be realized, the control of the relative position between a camera and the board is not needed, the requirement of an angle of shooting is reduced, the influence of a surrounding environment is small, a certain degreeof light changes can be borne, the use of large-scale machinery is not needed, the board can be analyzed through a mobile hone APP, thus the investment of equipment is saved, and the cumbersome task of board recognition and manual recording can be effectively reduced.

Description

technical field [0001] The invention relates to an identification method, in particular to an image-based Go score identification method. Background technique [0002] There are two main types of chess book recognition technologies. One is to use multiple frames of images for recognition, such as shooting continuous videos to judge where the chess piece falls through the difference between the frames before and after, or shooting an image of an empty chessboard in advance, and then placing it. After placing the chess pieces, take pictures of the chessboard, and use the difference between the two pictures to identify the chess score. The disadvantage of this method is that it needs to shoot multiple images, and the relative position and angle between the camera and the chessboard need to be fixed; another method can be Only a single picture can be used to identify the chess record, but such methods often require manual input of some information for identification. In additio...

AI Technical Summary

Problems solved by technology

[0002] There are two main types of existing chess record recognition technologies. One is to use multi-frame images for recognition, such as taking continuous videos to judge where the chess pieces fall through the frame difference between the front and back, or taking an image of an empty chessboard in advance, and then placing it After placing the chess pieces, take pictures of the chessboard, and use the difference between the two pictures to identify the chess record. The disadvantage of this method is that multiple images need to be taken, and the relative position and angle between the camera and the chessboard need to be fixed; another method can be Only a single picture can be used to identify chess records, but such methods often require manual input of some information for identification

In addition, these two methods have strict requirements on the surrounding environment and shooting angle of the chessboard, and they cannot automatically recognize the chess records on the existing chessboard in an open environment.

Images