Image Segmentation Based on a Two-Dimensional Histogram

Journal of Image and Graphics, 2015

Pattern Recognition, 1990

This paper describes a totally automatic non-parametric clustering algorithm and its application to unsupervised image segmentation. The clusters are found by mode analysis of the multidimensional histogram of the considered vectors through a non-iterative peak-climbing approach. Systematic methods for automatic selection of an appropriate histogram cell size are developed and discussed. The algorithm is easily parallelizable and is simulated on a SEQUENT parallel computer. Image segmentation is performed by clustering features extracted from small local areas of the image. Segmentation of textured, color, and gray-level images are considered. Eight-dimensional random field model based features, three-dimensional RGB components, and one-dimensional gray levels are utilized for these three types of images respectively. For texture segmentation, an image plane cluster validity procedure based on region growing of the mapped back clusters in the feature space is developed. Most of the phases are also parallelized resulting in almost linear speed ups. Quite satisfactory results are obtained in all cases.

In this paper, a new method for color image segmentation is presented. This method is based on histogram thresholding and correlation between the difference of color components. Hence, nearly all histogram thresholding methods work only in one or two dimensions of gray scale histogram, neighborhood, probability function or entropy. The proposed method will try to use color components as the main features of segmentation by finding the correlation between the peaks of histogram in each color component. It will help us to find main color components of each object and the background of image. While, we have main color components; it will be easy to use parallel processing to segment entire image at once without using any neighborhood window or losing any data in color space transform into gray scale. With these benefits, a fast and accurate method based on adaptive histogram thresholding is presented in this paper for segmentation of color images. The experimental results on benchmark datasets demonstrate the efficiency of the proposed method.

In this paper, a new color image segmentation method based on adaptive histogram thresholding and probabilistic clustering of random local samples is presented which is able to be used in parallel processing. Histogram thresholding methods are well known for their fast results and low time complexity, but nearly all thresholding methods work only over gray scale images and suffer from large deal of data loss during color space transformation to gray scale. In recent years, several approaches try to improve the accuracy of thresholding methods by adding the difference dimensions like entropy, probability function or pixel neighborhood to gray scale. Proposed method uses color channel components as main features of segmentation and then tries to find correlation between histogram's peaks and plains between each color components by using predefined random samples of image. This will be useful to define main color components of each object in image and then it will be easy to separate objects and background from each other with parallel processing. Low time complexity and ability of using parallel processing with high accuracy make this method a robust and reliable approach for color image processing in both RGB and HSI color spaces. The experimental results on benchmark datasets demonstrate that the proposed method is more efficient than the standard color image segmentation methods.

International Journal of Computer Applications, 2013

The prime objective of this paper is to implement an efficient improved color image segmentation method using local histogram and region merging technique. The goal of image segmentation is to cluster pixels into salient image regions, i.e. regions corresponding to individual surfaces, objects or natural parts of objects. Segmentation can be used for object recognition, occlusion boundary estimation within motion or stereo systems, image compression, image editing or image database look-up. Usually image segmentation is an initial and vital step in a series of processes aimed at overall image understanding. There are various techniques for image segmentation. In this research paper, a thorough work has done on the average local histogram of three different color spaces RGB, HSV & Lab. After that k-means clustering and labeling have done on the image for final segmentation.

Lecture Notes in Computer Science, 2011

An algorithm is presented to segment a color image based on the 3D histogram of colors. The peaks in the histogram, i.e., the connected components of colors with locally maximal occurrence, are detected. Each peak is associated a representative color, which is the color of the centroid of the peak. Peaks are processed in decreasing occurrence order, starting from the peak with the maximal occurrence, with the purpose of maintaining only the representative colors corresponding to the dominant peaks. To this aim, each analyzed peak groups under its representative color those colors, present in the histogram and that have not been grouped to any already analyzed peak, such that their distance from the centroid of the peak is smaller than a priori fixed value. At the end of the grouping process, a number of representative colors, generally substantially smaller than the number of initial peaks, is obtained, which are used to identify the regions into which the color image is segmented. Since the histogram does not take into account spatial information, the image is likely to result over-segmented and a merging step, based on the size of the segmentation regions, is performed to reduce this drawback.

— This study is focused on histogram thresholding methods automatically. In computer vision, Image segmentation is an initial and vital step in a series of processes aimed at overall image understanding. In other words Segmentation refers to the process of partitioning a digital image into the multiple segments (set of pixels as known as super pixels). Two very simple image segmentation techniques that are based on the gray level histogram of an image are Thresholding and Clustering. Thresholding method is widely used for image segmentation approach. It is useful in discriminating foreground from the background. By selecting an adequate threshold value T, or automatically computing threshold value T, the gray level image can be converted in to binary image. Several methods are there to find the threshold automatically for image segmentation. Some of the methods like Otsu, Kapur, Triangle, Iterative and also manually threshold is calculated for different type of images like X-ray computed tomography (CT-Scan), magnetic resonance imaging (MRI), synthetic aperture radar (SAR), Ultrasound image were explained and the results are presented to show the validity of the methods.

Pattern Recognition Letters, 2006

This paper proposes a new method of image thresholding by using cluster organization from the histogram of an image. A new similarity measure proposed is based on inter-class variance of the clusters to be merged and the intra-class variance of the new merged cluster. Experiments on practical images illustrate the effectiveness of the new method.

Image segmentation is often used to distinguish the foreground from the background. This paper proposes a novel method of image thresholding using the optimal histogram segmentation by the cluster organization based on the similarity between adjacent clusters. Since this method is not based on the minimization of a function, the problem of selecting the threshold at the local minima is avoided. This approach overcomes the local minima that affect most of the conventional methods by maximizing the between-class and minimizing within-class objects. Agglomerative clustering is used in this method so as to merge two adjacent clusters in the histogram. The distance measurement using discriminant analysis is adapted from the criterion function defined by Otsu. It directly approaches the feasibility of evaluating the goodness of every pair and automatically grouping the closest pair. The most similar pair is selected, which is the most homogeneous one. In addition, this pair should be the ...

– Image segmentation is one of the key steps in many problem solving methods in image processing applications. This paper addresses the image segmentation problem for content-based image retrieval applications. Our method is based on multi-thresholding histogram in CIE L*a*b* color space and we have used average filter to discover the peaks of histogram. An algorithm similar to morphology is used to merge the initial segments. We have used the Berkeley dataset to evaluate our method and our result is compared with the results of methods provided by the Berkeley University.