(OpenCV/C++) Image Rectitication/Transformation
Hi, am new to computer vision, began Monday, and need to perform a perspective transformation in images with these characteristics:
... To look like this:
Researched in various forums, but still did not get results, I heard that the characteristics of a chessboard help in rectifying / transformation of the image. Would appreciate it very much if someone help me at least with step by step. Thank U!
Have a look at this: calibration
Thanks, but i think that this solution not fits on my problem, cause i'm working with a image and this solution is with a camera calibration, with video. Or I can be mistaken. However thanks again!
is the number of squares of the chessboard considered as known or not, because then you can use the findchessboardcorners() function in combination with this here in order to detect the chessboard easier. Moreover, your input image is distorted, you will need to undistort it first in order to obtain better results.
Thank you Theodore. I will use a standard chess board, because I will shoot a subject for analysis, on this board, and then perform the transformation on the image, right after this, perform the analysis with the object in a better position.
Theodore, in this example you gave me, you must find the 4 corners of the chess board, but with the findchessboardcorners are not returned the 4 corners, but is returned something like this: https://uploaddeimagens.com.br/imagen...
obtaining these corners you can find the four ones that are needed. Moreover, if you want to use a more automatic approach you might try to implement the solution proposed here.