DTM: Deformable Template Matching

TL;DR

This paper introduces Deformable Template Matching (DTM), a new non-training-based template matching method that uses predefined deformation rules, improving performance in object recognition and feature matching tasks.

Contribution

The paper presents DTM, a novel deformable template matching algorithm that does not require training and applies deformation rules for improved matching accuracy.

Findings

DTM outperforms conventional template matching algorithms on PASCAL VOC 07.

Applying DTM to SIFT features increases the number of matched features.

Deformable matching enhances robustness in object recognition and feature matching.

Abstract

A novel template matching algorithm that can incorporate the concept of deformable parts, is presented in this paper. Unlike the deformable part model (DPM) employed in object recognition, the proposed template-matching approach called Deformable Template Matching (DTM) does not require a training step. Instead, deformation is achieved by a set of predefined basic rules (e.g. the left sub-patch cannot pass across the right patch). Experimental evaluation of this new method using the PASCAL VOC 07 dataset demonstrated substantial performance improvement over conventional template matching algorithms. Additionally, to confirm the applicability of DTM, the concept is applied to the generation of a rotation-invariant SIFT descriptor. Experimental evaluation employing deformable matching of SIFT features shows an increased number of matching features compared to a conventional SIFT matching.