A Dual-Bound Algorithm for Very Fast and Exact Template Matching

Abstract

Recently proposed fast template matching techniques employ rejection schemes derived from lower bounds on the match measure. This paper generalizes that idea and shows that in addition to lower bounds, upper bounds on the match measure can be used to accelerate the search. An algorithm is proposed that utilizes both lower and upper bounds to detect the k best matches in an image. The performance of this dual-bound algorithm is guaranteed; it always detects the k best matches. Theoretical analysis and experimental results show that its runtime compares favorably with previously proposed real-time exact template-matching schemes.

Summarization-Based Image Resizing by Intelligent Object Carving

“…The matching measure involved is simple, yet more effective than some "more accurate" template matching algorithms. For instance, the dual-bound [32] algorithm often fails to detect objects which are relatively dissimilar with the template, as shown in Fig. 5d.…”

“…5 shows the importance of the outer boundary enhancement in our algorithm. We can detect the balloons with strongly different inner textures accurately, whereas RepFinder cannot, for the BBM in their algorithm is constructed by calculating the outer and inner boundaries [9], MSCR [35] and dual-bound [32] (k ¼ 10), our result is more accurate. The image on the left shows the template and its shape feature points.…”

Pattern Matching Using Partial Fourier Spectrum

“…Also, use of lower bound function can be generalized to use of simultaneous lower and upper bound functions [23] which are determined by the first k sub-images with least dissimilarity.…”

A template extraction approach for image recognition

“…In the two dimensional scatter image matching application, the information is analyzed which about the features and pixel data. Especially in the grayscale image matching process, when the target image is added salt and pepper noise, it is able to improve matching accuracy effectively [4], which is used by weighted evaluation mechanism on the basis of the existing noise immunity.…”