For any chess board pictures, can quickly high identification accuracy, and saved as a standard format, what is design thinking? I didn't want to please the great god give directions?

CodePudding user response:

What's the idea?

CodePudding user response:

What chess, go? Is black and white

Generally the flags pictures are small, at most 64 * 64, use TMemoryStream loading images, the flags with CompareMem memory than have to
Of course can also be pixel by contrast

CodePudding user response:

Go and have a code? consult

CodePudding user response:

If it is go fuck ah, also directly determine the center pieces of a pixel, the pixel is black is black, on the other hand,

CodePudding user response:

This should be a complex technology, although some simpler than the fingerprint and face recognition, but more complicated than the verification code identification

CodePudding user response:

LZ should be to make a video game board, rather than real chessboard, taken with a camera
Go move later in video games are polite reticle is central, get the board can determine the size of the pieces of grid lines

CodePudding user response:

Ha ha, I just wrote a before the Spring Festival, I also take a name call "screenshots continue to implement", you can download a try:
http://download.csdn.net/detail/bluesen/6905157

CodePudding user response:

reference 4 floor sololie response:

if it is go fuck ah, also directly determine the center of a pixel pieces, the pixel is black is black, whereas

Complicated than imagination, some of them are shows the digital move, some of the pieces is stereo images,

CodePudding user response:

Want to download the LS learning resources under reference, but a look at five points, not under the
Best is 0, download a will also get 1 minute, and that there will be more people download, fan is the limit, 100 to 100 points to download you won't get scores, the set is cheaper to CSDN download points to let them make money, don't need to,

CodePudding user response:

To go, my idea is to the entire board, binary image, and then find out the first grid grid, the grid out can know the size of the pieces and the size of the chessboard,
Move later is located in the center of the graticule mesh, so also can locate the area of each move later, find out the size after erasing grid lines from the original image, behind facilitate analysis pieces,
As to pawn is stereo images, pure color, or has the form actually don't have to care about, such as reflective scanning the entire board pictures, go there are only two pieces, find out the regularity of the same size pieces, such as 64 * 64 pixels group, plus a is no move later area, only three images,
Distinguish the spots from white, is to see every piece in this set of pixels is much more than black or white, if it is pure white is no pieces,

Maybe so, practice.

CodePudding user response:

references 9 f sololie response:

want to download the LS learning resources under reference, but a look at five points, not under the
Best is 0, download a will also get 1 minute, and that there will be more people download, fan is the limit, 100 to 100 points to download you won't get scores, the set is cheaper to CSDN download points to let them make money, don't need to,

To listen to you, don't points now, download address is:
http://download.csdn.net/detail/bluesen/6905691

CodePudding user response:

How to just the software, not the source code,,,,

CodePudding user response:

The

refer to 12 floor sololie reply:

how just software, is not the source code,,,,

The principle of software also can consult, besides can be discussion, also not to the source code,

CodePudding user response:

CodePudding user response:

Pattern recognition
Machine vision
Artificial intelligence (ai)
Now in vogue deep learning algorithms, belongs to a kind of neural network algorithm,

CodePudding user response:

Demand is not obvious, what main used for?

CodePudding user response:

Should be similar to the software I wrote, that is, the current position (such as ground city, Tom go, China game center and other game software) screenshots, identify the location of the black and white pieces, into your own application, you can try next,
Equivalent to j software import is not move, but a picture,

Such as the software (http://download.csdn.net/detail/bluesen/6905691) function description I wrote is like this:
1) in ground city (or other game software) game, is not clear, need to study real-time game, this time will be the game screen screenshots and then click the start button of the software can begin to try, after

2). In group discussions with fellow chess player, analysis of the situation, will the game screen screenshots and then click the start button of the software, select "mark", click on the possible source, the software will automatically labeled "A", "B", "C", "D" and so on, then you will screen screenshots of the software into the group and ask everyone under which points,

3). To help the inside of the group of fellow chess player, the fellow chess player's shot into the form of this software, you can according to their own ideas to try, then the software screen screenshots on the group,

4). J, j, chess book