Template matching for noisy images
Sergey Shaydullin, Andrei Turkin, Nikolay Chelyshev · 2015
In many applications of computer vision one must deal with images that are distorted by noise which level cannot be estimated exactly. In this work we discuss a template matching framework which uses optimal nonlinear distortion-tolerant filtering technique and can be applied for noisy images. In this paper we assess the accuracy of this framework under conditions of uncertainty in level of noise.