Affine template matching by differential evolution with adaptive two‐part search

Abstract

In this paper, we address the affine template matching of general images. The extensive search space of affine transformations necessitates effective searches of the global optimum. The proposed method utilizes differential evolution (DE), which is a method of metaheuristic optimization, to achieve that goal. Self-adaptive DEs can be useful and are applicable in a wide range of studies as they tune crossover rate and scaling factor (F) themselves over generation iteration. However, this approach is not particularly good for affine template matching because the population often converges to local optima. In order to solve this problem, the population is divided into two equal groups for exploitation and exploration. The former group utilizes current-to-best/1, and the latter group adopts improved current-to-rand/1 for the mutation scheme. Furthermore, the proportion of the population sizes of the two groups are linearly changed on the basis of the best sum of absolute difference error measurements over each generation. These ideas are easy and simple, but experimental results have revealed our method to be more accurate than the state-of-the-art method.

Perspective template matching by differential evolution introducing the group remaking

“…In order to solve the problem of Zhang's GA, Sato et al proposed a method using DE (Differential Evolution) [5]. Since the population size is fixed similar to general evolutionary computation methods, memory size problem can be avoided.…”

“…Next, DE with adaptive two-part search [5] to be focused in this research is explained. Before the detailed explanation of the algorithm, a brief summary is firstly described here.…”

“…The reason why this scheme is introduced is the local search should be weighted relatively because f g,best close to 1.0 means the population converges to the global optimum. By changing the both group sizes adaptively, this DE can achieve better performance than Zhang's GA on affine template matching [5].…”

“…In order to overcome this problem, Sato et al utilized differential a Correspondence to: Junya Sato. E-mail: jsato@gifu-u.ac.jp *Faculty of Engineering, Gifu University, 1-1, Yanagido, Gifu-shi, Gifu, Japan, 501-1193 **Faculty of Science and Engineering, Iwate University, 4-3-5, Ueda, Morioka, Iwate, Japan, 020-8551 evolution (DE) [5]. This method does not necessitate the large population size.…”

“…The search ranges of each parameter are described in Table I. Considering comparative experiments, the search ranges are the same to the related work [4,5].…”