Rifat Kurban

Fusion of multi-focus images using differential evolution algorithm

In an image captured by a CCD/CMOS sensor with an optical lens, only the objects within the depth of field are sharply focused. One way to get an everywhere-in-focus image is to fuse the images of the same scene which are taken with different focal settings. In this paper, a novel optimal method for multi-focus image fusion using differential evolution algorithm is presented. The source images are first decomposed into blocks. Then, the sharper blocks are selected by employing a sharpness criterion function. The selected blocks are finally combined to construct the fused image. The motivation of the proposed method lies in the fact that an optimized block size could be more effective than a fixed block size. Experimental results show that the proposed method outperforms other traditional methods and genetic algorithm based method in terms of both quantitative and visual evaluations.

Comparison of evolutionary and swarm based computational techniques for multilevel color image thresholding

This paper introduces the comparison of evolutionary and swarm-based optimization algorithms for multilevel color image thresholding problem which is a process used for segmentation of an image into different regions. Thresholding has various applications such as video image compression, geovideo and document processing, particle counting, and object recognition. Evolutionary and swarm-based computation techniques are widely used to reduce the computational complexity of the multilevel thresholding problem. In this study, well-known evolutionary algorithms such as Evolution Strategy, Genetic Algorithm, Differential Evolution, Adaptive Differential Evolution and swarm-based algorithms such as Particle Swarm Optimization, Artificial Bee Colony, Cuckoo Search and Differential Search Algorithm have been used for solving multilevel thresholding problem. Kapurs entropy is used as the fitness function to be maximized. Experiments are conducted on 20 different test images to compare the algorithms in terms of quality, running CPU times and compression ratios. According to the statistical analysis of objective values, swarm based algorithms are more accurate and robust than evolutionary algorithms in general. However, experimental results exposed that evolutionary algorithms are faster than swarm based algorithms in terms of CPU running times.

New optimised region-based multi-scale image fusion method for thermal and visible images

On constructing a fused image by employing only individual pixels or a set of pixels within a small neighbourhood of the images (SIs) acquired from the same scene, pixel-based fusion techniques suffer from some drawbacks, such as blurring effects, high sensitivity to noise and misregistration. To overcome these drawbacks, this study proposes a new region-based image fusion method for thermal and visible images. Since different regions with certain properties need to be emphasised differently in the fused image, the corresponding regions of the SIs are optimally merged to obtain the fused image by employing multiple weighting factors (WFs). To improve the quality of the fused images, WFs were optimised by employing the differential evolution algorithm. Furthermore, a new quality metric was also developed to measure the quality of the fused images during the optimisation process. Experimental results show the feasibility of the proposed method.

Evaluation of criterion functions for the fusion of multi-focus noisy images

Digital images are distorted by impulsive noise (IN) which caused by image sensors and communication channels in many practical applications. This noise may cause miscalculation of sharpness values in image fusion applications. In this paper, traditional criterion functions and frequency selective median filter (FSWM) based criterion function are evaluated for the fusion of multi-focus images in the presence of IN. Experimental results show that the FSWM filter is better than the other focus measures.

Extending depth-of-field of a digital camera using particle swarm optimization based image fusion

Images obtained by an optical camera are seriously affected from the limited depth of the field issue. That is, the objects located in-front-of or behind the focus of the camera are blurred. Everywhere-in-focus images can be obtained by combining focused regions of images which were captured by different focal settings. In this paper, an optimal region based spatial domain image fusion approach based on selecting sharper regions from the multi-focus source images is proposed. Size of the block type regions are optimized by using particle swarm optimization (PSO) algorithm. Quantitative and subjective experimental results of different image sets show that proposed method is better than traditional wavelet and Laplacian pyramid based and also genetic algorithm (GA) based methods.

Extending depth-of-field by image fusion using multi-objective genetic algorithm

Optical imaging systems used in industrial vision applications suffer from the limited depth of field problem. Image fusion is a practical way to obtain an everywhere-in-focus image of a scene. In this paper, a spatial domain image fusion method using multi-objective genetic algorithm is proposed to obtain an optimal fused image. Experimental results conducted with industrial images show that the proposed method outperforms Laplacian Pyramid and DWT methods.