Soft Computing Approaches for Image Segmentation

Image Segmentation

As the computer technology is growing day by day the inference of images in daily life has been seen quite often. The tremendous growth of internet has also incorporate use of images. With the number of applications such as social networking, banking, graphical password and so on the images are used everywhere thus incorporates image processing. The images present around us are digital in nature and thus processed by using computer or other smart devices compatible with computers. So when a digital image has been processed by computer to obtain some useful information, enhance image quality, manipulate or analyze something from the given image is known as image processing or simply it can be stated that converting an image into the digital form for getting some information from it.

Figure 1.

Image Processing

To incorporate images and process it through the computer becomes a challenging and most important task as the frequency of images are used growing rapidly. The most important task in an image is to detect what it contains for example consider an image shown in figure 2 let us suppose that a computer is given with the same image and a man is given with the same image to recognize the contents of the image, the man by looking at the image can easily find out that the image contains fruits but computer cannot recognize it so not only making image processed through the computer system to be in digital manner but also make the processing meaningful by detecting the contents is also a major issue. Convenient application of image segmentation range from medical application (locate tumors, breast cancer, computer guided surgery, treatment planning, measure tissue volumes etc), locate objects in satellite images (water, forest, roads, buildings, etc), Face recognition, Finger print recognition, Plant diseases recognition etc.

Image segmentation or object extraction is the method by which the contents of a digital image are recognized by the computer. Image segmentation is the process of breaking the image into very small parts so that the objects can be identified and some useful information can be generated. So in the given image figure 2 when its segmentation has been done than it could recognize the objects and thus stated that the image given to the computer contains fruit.

Figure 2.

Object extraction

Object extraction (locate, identify, boundaries detection) from the image is the vital topic in the field of digital image processing. The main purpose of image segmentation is to segregate the image into necessary regions with respect to the appropriate locations which are visually distinct and uniform in accordance to some property such as texture, colour or gray level so, as to make image into something that is more meaningful and easier to analyze. Moreover segmentation of an image depends on the problem domain or application. Image segmentation can be defined as follows:

or