Abstract
Multichannel images, i.e., images of the same object or scene taken in different spectral bands or with different imaging modalities/settings, are common in many applications. For example, multispectral images contain several wavelength bands and hence, have richer information than color images. Multichannel magnetic resonance imaging and multichannel computed tomography images are common in medical imaging diagnostics, and multimodal images are also routinely used in art investigation. All the methods for grayscale images can be applied to multichannel images by processing each channel/band separately. However, it requires vast computational time, especially for the task of searching for overlapping patches similar to a given query patch. To address this problem, we propose a three-dimensional orthonormal tree-structured Haar transform (3D-OTSHT) targeting fast full search equivalent for three-dimensional block matching in multichannel images. The use of a three-dimensional integral image significantly saves time to obtain the 3D-OTSHT coefficients. We demonstrate superior performance of the proposed block matching.
Highlights
Block matching is a fundamental tool used to search for blocks similar to a given query.It has been widely used in solving various image processing problems, like object recognition and tracking [1], image registration [2], image analysis [3], image restoration [4], to name a few
We propose here fast full search (FS)-equivalent three-dimensional (3D) block matching for multichannel image using three-dimensional orthonormal tree-structured Haar transform (3D-OTSHT)
We performed 3D-OTSHT for three-dimensional (3D) block matching in order to evaluate the pruning performance and elapsed time of the proposed method using multichannel images
Summary
Block matching is a fundamental tool used to search for blocks (patches) similar to a given query. It has been widely used in solving various image processing problems, like object recognition and tracking [1], image registration [2], image analysis [3], image restoration [4], to name a few. Block matching searches for patches of the same size as a query and is sensitive to deformation. In this sense, it is different from some common image descriptors such as scale-invariant feature transform (SIFT) [5]. The larger the image size and number of image bands, the harder it is to use FS
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have