This chapter proposes an unsupervised grayscale image segmentation method based on the Firefly and Artificial Bee Colony algorithms. The Firefly Algorithm is applied in a histogram-based research of cluster centroids to determine the number of clusters and the gray levels, successively used in the initialization step for the parameter estimation of a Gaussian Mixture Model. The coefficients of the linear super-position of Gaussians can be thought of as prior probabilities of each component. Applying the Bayes rule, the posterior probabilities of the grayscale intensities are evaluated and their maxima are used to assign each pixel to clusters. Subsequently, region spatial information is extracted to form homogeneous regions through ABC algorithm. Initially, scout bees are moving on the search space describing random paths, with food sources given by the detected homogeneous regions. Then onlooker bees rush to scouts' aid proportionally to unclassified pixels enclosed into the bounded boxes of the discovered regions.
TopIntroduction
Firefly Algorithm and Gaussian Mixture Model
Image segmentation is the process of partitioning a digital image into multiple segments of pixels. The goal of image segmentation is to simplify the representation of an image in one more meaningful and easier to understand. It is often an essential step in image analysis, object representation, visualization, and many other image processing issues and it is aimed at facilitating the tasks at higher levels, such as object detection and recognition. Image segmentation is a preliminary step in image processing, playing a role of great relevance in object recognition and classification, machine learning, image learning and understanding (Gonzalez & Wood, 2007).
In this work, we propose a method for grayscale image segmentation and edge extraction that is composed of two different phases in which are applied two bio-inspired algorithms (Osuna-Enciso, 2014) Firefly and the Artificial Bee Colony algorithms. In the first phase, a pixel-based segmentation is performed applying the FA algorithm and the Gaussian Mixture Model. Classification of pixels in different gray levels is obtained through a histogram-based segmentation approach. To do so, we utilize the nature-inspired FA algorithm for defining automatically the number of clusters given by histogram maxima.
Artificial Bee Colony Algorithm
After assigning all pixels to the corresponding gray classes, we carry out the second step with a region-based segmentation. For this purpose, the ABC algorithm was performed to complete the segmentation process, which resulted greatly simplified since there are only a few levels of gray. The region growing technique is implemented using randomly generated scout bees as initial seed points. The scout bees are moving on the search space, i.e. the gray image, describing random paths. The food sources are defined by homogeneous regions. Once a region of a specific gray intensity has been found, the scout bee comes back to the hive for executing the waggle dance in order to involve onlooker bees in the exploitation phase. Then onlooker bees give rise to a local search, rushing to scouts’ aid proportionally to the size the of the bounded boxes of regions discovered by scout bees and to the number of unclassified pixels included into them.
Background
The basic goal of any image segmentation process is to subdivide an image into components belonging to different objects or to different parts of an object. Theoretically, pixels derived by the same component should have similar intensities, forming a connected region (Russ & Neal, 2015). During the last decades many segmentation methods have been proposed in literature, for an in-depth overview of clustering techniques refer to (Nikhil & Sankar,1993; Jain, Murry, & Flynn., 1999; Herbert & Pantofaru, 2005; Zaitoun & Musbah, 2015). The segmentation techniques are categorized into three main classes: pixel-based, region-based and edge-based schemes (Haralick & Shapiro, 1985). The threshold-based segmentation approach belongs to the first category, it is one of the simplest and widely used in computer vision. In thresholding methods, pixels are partitioned depending on their intensity levels, recurring to global thresholding, with a single threshold T, or to variable thresholding if T changes over the image, or to multiple thresholding values T1, …Tn. Global thresholding maps a gray-valued image into a binary image, hence it can be suitable to segment an image into objects and background. Threshold segmentation can be extended to use multiple thresholds to decompose an image into more than two segments. Many methods exist to select threshold values for a segmentation task, among them we could mention the image histogram-based approaches, which are often valuable tools in establishing suitable thresholding values (Ridler & Calvard, 1978; Chen & Chen, 2009).
Region-based segmentation algorithms operate iteratively by grouping together pixels that have neighbours with similar values and splitting groups of pixels that are dissimilar.