How to realize Cylindrical projection by OpenCV in C language
It is believed that many inexperienced people don't know what to do about how to realize cylindrical projection in OpenCV in C language. therefore, this paper summarizes the causes and solutions of the problem. I hope you can solve this problem through this article.
Preface
When doing panoramic stitching, in order to maintain the consistency of spatial constraints and vision in the picture, it is necessary to carry out cylindrical projection, otherwise the more distant from the central image, the greater the deformation after stitching.
The cylindrical projection formula is
The implementation code is for color images int main () {cv::Mat image1 = cv::imread ("images/1.jpg", 1); if (! image1.data) return 0; imshow ("image1", image1); Mat imgOut = Mat (image1.rows, image1.cols, CV_8UC3); float w = image1.cols; float h = image1.rows Float f = (w / 2) / atan (PI / 8); for (int I = 0; I
< image1.rows; i++) { for (int j = 0; j < image1.cols; j++) { float x = j; float y = i; float x1 = f * atan((x - w / 2) / f) + f * atan(w / (2.0f * f)); float y1 = f * (y - h / 2.0f) / sqrt((x - w / 2.0f) * (x - w / 2.0f) + f * f) + h / 2.0f; int col = (int)(x1 + 0.5f);//加0.5是为了四舍五入 int row = (int)(y1 + 0.5f);//加0.5是为了四舍五入 if (col < image1.cols && row < image1.rows) { imgOut.at(row, col)[0] = image1.at(i, j)[0]; imgOut.at(row, col)[1] = image1.at(i, j)[1]; imgOut.at(row, col)[2] = image1.at(i, j)[2]; } } } imshow("imgOut", imgOut); waitKey(0); return 0;} 实现效果
For grayscale images cv::Mat image1 = cv::imread ("E:\ zcb_work\ 2113\ pic2\ k.jpg", 0); if (! image1.data) return 0; imshow ("image1", image1); cv::Mat image2 = cv::imread ("E:\ zcb_work\ 2113\ pic2\ j.jpg", 0) If (! image2.data) return 0; imshow ("image2", image2); Mat imgOut1 = Mat (image1.rows, image1.cols, CV_8UC1); imgOut1.setTo (0); Mat imgOut2 = Mat (image2.rows, image2.cols, CV_8UC1); imgOut2.setTo (0); float w = image1.cols; float h = image1.rows Float f = (w / 2) / atan (PI / 8); for (int I = 0; I < image1.rows; iTunes +) {for (int j = 0; j < image1.cols; jacks +) {float x = j; float y = I Float x1 = f * atan ((x-w / 2) / f) + f * atan (w / (2.0f * f)); float Y1 = f * (y-h / 2.0f) / sqrt ((x-w / 2.0f) * (x-w / 2.0f) + f * f) + h / 2.0f Int col = (int) (x1 + 0.5f); / / add 0.5 to round int row = (int) (y1 + 0.5f) / / add 0.5 to round if (col < image1.cols & & row < image1.rows) {imgOut1.at (row, col) = image1.at (I, j); imgOut2.at (row, col) = image2.at (I, j) / / imgOut.at (row, col) [1] = image1.at (I, j) [1]; / / imgOut.at (row, col) [2] = image1.at (I, j) [2];} imshow ("imgOut1", imgOut1) Imshow ("imgOut2", imgOut2)
Realize the effect
Original drawing
Cylindrical projection
Using surf algorithm, feature detection
Synthesize it like this, hehe
After reading the above, have you mastered how to achieve cylindrical projection with OpenCV in C language? If you want to learn more skills or want to know more about it, you are welcome to follow the industry information channel, thank you for reading!
