Image Enhancement Techniques Using Particle Swarm Optimization Technique

Introduction

In real life scenario, huge amount of data available in the form of images and its quality need to be enhanced for making it ready to apply high level image processing techniques like pattern recognition, feature extraction, analysis and any other automation engineering applications. In general, process of improving the quality of images is called image enhancement and it is considered as one of the important preprocessing techniques for high level processing of images. The enhancement can be performed by hardware devices through software algorithms but this chapter deals with various software techniques in particular. Gorai and Ghosh (2009) outlined in their work, that in image enhancement process images are transferred from one level to another level through the process of intensity transformation or filtering approaches to improve the perception of information for human viewers, or to provide better input for other automated engineering applications.

In general image enhancement processes include histogram equalization, intensity transformation, smoothing, sharpening, contrast adjustment, frequency domain filtering and pseudo coloring as explained by Gonzales and Woods (1987). Majority of the image enhancement work usually manipulates the image histogram by some transformation function to obtain the required contrast enhancement. Consequently, this operation also delivers the maximum information contained in the image. In detail, the histogram transformation is considered as one of the fundamental processes for contrast enhancement of gray level images, which facilitates the subsequent higher level operations such as detection and identification. Histogram processing could be carried out either by local processing or by global processing. In linear contrast stretching, the gray value of input image is mapped to the complete range of gray values in the output image. Similarly in Pseudo coloring gray values are mapped to color values artificially to highlight and enhance the important regions in images. Generalizing gray scale image enhancement to color image enhancement is not a trivial task. Several factors, such as selection of a color model, characteristics of the human visual system, and color contrast sensitivity, must be considered for color image enhancement. Color images could be separated into the chromaticity and intensity components. The intensity components are also called luminance components. The quality of color images could be increased by applying any of the image enhancement techniques on luminance part of color images.

Some of the image/signal processing applications include iris recognition, fruit quality grading, face detection, object recognition, image segmentation, synthetic aperture radar image processing, image classification, image fusion and many more. Recently various areas of automation engineering find its use and contribution of image processing techniques and it includes medical imaging, people identification proposed by Xiaodong and Zhong (2006), indoor security surveillance elaborated by Xiang and Yan (2007) and crowd monitoring in outdoor environments proposed by Dorigo and Stützle (2004) etc. Many classical enhancement methods exist but it does meet the requirements of automation engineering.