Uploaded January 2015 | Updated September 2026, 2 weeks ago
The basic form of template matching is pretty boring and is not very robust. But more robust approaches such as detecting keypoints, extracting local invariant descriptors, and performing keypoint matching using RANSAC and LMEDS can be overkill.
In this blog post I'll show you a method to handle when you need a more powerful method than simple template matching, but don't want the expense of keypoint matching.
This approach is called "multi-scale template matching" and can be used to find objects in images.
Click here for the full blog post: pyimagesearch.com/2015/01/26/multi-scale-template-matching-using-python-opencv
The basic form of template matching is pretty boring and is not very robust. But more robust approaches such as detecting keypoints, extracting local invariant descriptors, and performing keypoint matching using RANSAC and LMEDS can be overkill.
In this blog post I'll show you a method to handle when you need a more powerful method than simple template matching, but don't want the expense of keypoint matching.
This approach is called "multi-scale template matching" and can be used to find objects in images.
Click here for the full blog post: pyimagesearch.com/2015/01/26/multi-scale-template-matching-using-python-opencv
