zelianwen / image-registration Goto Github PK

感谢作者大大提供的代码，让我这个初学菜鸡能够看到运行结果，去知网拜读了一下您的论文，写的非常好！！再次感谢

I note if I change all CV_32FC1 in your code into CV_16SC1, then the programe can save a lot memory consumption. But the sift_1.detect will fail to get that keypoints_1 in following. How to change the other code for the change?

如下的一个LMS函数代码中,如果当match1_xy.rows等于match2_xy.rows 等于0 的时候，会在solve(A, B, values, DECOMP_QR);函数求解逆矩阵的时候报错，需在下面增加一个排除条件
if (match1_xy.rows ==0 || match2_xy.rows==0)
CV_Error(Error::StsBadArg, "LMS模块输入特征点为0！");

也可以在LMS(consensus_arr1, consensus_arr2, model, rmse);进来前判定arr1和arr2的rows是否为空，如果为空则给出提示信息并直接return。

static Mat LMS(const Mat&match1_xy, const Mat &match2_xy, string model, float &rmse)
{
if (match1_xy.rows != match2_xy.rows)
CV_Error(Error::StsBadArg, "LMS模块输入特征点对个数不一致！");

.
.
.

}

您好，
我最近也在做图像配准的工作，非常感谢您分享自己的工作。但是我在跑您的代码的时候，发现sift-opencv249和sift_static.exe的结果并不一致，所以想问您，sift_static.exe又增加了什么优化工作，或者您是否愿意上传对应的新的sift-opencv249代码，非常感谢！
小池

Thanks for sharing the repository!

When I read the code,I found that there is only the main angle used in the function “calc_sift_descriptor.m”.
Do you have any plan in the future how to use the second highest orientation?
Although only about 15% of points are assigned multiple orientations, but these contribute significantly to the stability of matching.

Thanks again!

The code written is very good.It is very helpful for the beginners.Thank you so much for your sharing.

Hi ,
I am new to SAR technology. I am working on coregistration(Terrasarx images) using SIFT/AKAZE feature extraction techniques. I extracted matched key points from master and slave images and also calculated homography/RANSAC.
How to calculate phase values? I also want to do measure the error (pixel level) and compared to Geofloki results?
can anyone help me in this regard?