Colour image segmentation in various illumination circumstances

1991

Because segmentation results of grey-value images might be negatively a ected by the presence of intensity changes due to shadows and surface curvature in those images, we have investigated to what extent these segmentation results can be improved by using color information. The in uence of di erent color systems has been examined as well as two segmentation methods were investigated: a clustering and a region growing technique. The clustering technique is based on the k-means algorithm, and is iterative. The region growing technique is derived from the split&merge algorithm proposed by 4].In the region growing technique, homogeneity criteria play an important role. Two homogeneity criteria have been implemented: a homogeneity criterion based on functional approximation and the mean&variance homogeneity criterion. The results show a reasonable insensibility to changes in intensity due to shadows and surface curvature, whereby the region growing technique produces better results than the clustering algorithm. Color systems which are linear combinations of the same RGB basis provide similar segmentation results. The IHS, rgb and the L a b color systems provide poor segmentation results when intensity is small. The rgb color system provides the best segmentation results if the image contains intensity changes.

in this work, a new approach to fully automatic color image segmentation, called JSEG, is presented. First, colors in the image are quantized to several representing classes that can be used to differentiate regions in the image. Then, image pixel colors are replaced by their corresponding color class labels, thus forming a class-map of the image. A criterion for “good” segmentation using this class-map is proposed. Applying the criterion to local windows in the class-map results in the “J-image”, in which high and low values correspond to possible region boundaries and region centers, respectively. A region growing method is then used to segment the image based on the multi-scale J-images. Experiments show that JSEG provides good segmentation results on a variety of images.

Image Segmentation plays a vital role in image processing. The research in this area is still relevant due to its wide applications. Image segmentation is a process of assigning a label to every pixel in an image such that pixels with same label share certain visual characteristics. Sometimes it becomes necessary to calculate the total number of colors from the given RGB image to quantize the image, to detect cancer and brain tumour. The goal of this paper is to provide the best algorithm for image segmentation.

Eleventh International Conference on Machine Vision (ICMV 2018)

In this work we discuss the known algorithms for linear colour segmentation based on a physical approach and propose a new modification of segmentation algorithm. This algorithm is based on a region adjacency graph framework without a pre-segmentation stage. Proposed edge weight functions are defined from linear image model with normal noise. The colour space projective transform is introduced as a novel pre-processing technique for better handling of shadow and highlight areas. The resulting algorithm is tested on a benchmark dataset consisting of the images of 19 natural scenes selected from the Barnard's DXC-930 SFU dataset and 12 natural scene images newly published for common use. The dataset is provided with pixel-by-pixel ground truth colour segmentation for every image. Using this dataset, we show that the proposed algorithm modifications lead to qualitative advantages over other model-based segmentation algorithms, and also show the positive effect of each proposed modification. The source code and datasets for this work are available for free access at http://github.com/visillect/segmentation.

1994

Image segmentation, i.e., identi#cation of homogeneous regions in the image, hasbeen the subject of considerable research activityover the last three decades. Manyalgorithms have been elaborated for gray scale images. However, the problem ofsegmentation for colour images, which convey much more information about objectsin scenes, has received much less attention of scienti#c community. While severalsurveys of monochrome image segmentation techniques were

2019

In this work we discuss the known algorithms for linear colour segmentation based on a physical approach and propose a new modification of segmentation algorithm. This algorithm is based on a region adjacency graph framework without a pre-segmentation stage. Proposed edge weight functions are defined from linear image model with normal noise. The colour space projective transform is introduced as a novel pre-processing technique for better handling of shadow and highlight areas. The resulting algorithm is tested on a benchmark dataset consisting of the images of 19 natural scenes selected from the Barnard’s DXC-930 SFU dataset and 12 natural scene images newly published for common use. The dataset is provided with pixel-by-pixel ground truth colour segmentation for every image. Using this dataset, we show that the proposed algorithm modifications lead to qualitative advantages over other model-based segmentation algorithms, and also show the positive effect of each proposed modifica...

Machine Vision and Applications, 2011

The aim of this paper is to propose a new methodology for color image segmentation. We have developed an image processing technique, based on color mixture, considering how painters do to overlap layers of various hues of paint on creating oil paintings. We also have evaluated the distribution of cones in the human retina for the interpretation of these colors, and we have proposed a schema for the color mixture weight. This method expresses the mixture of black, blue, green, cyan, red, magenta, yellow and white colors quantified by the binary weight of the color that makes up the pixels of an RGB image with 8 bits per channel. The color mixture generates planes that intersect the RGB cube, defining the HSM (Hue,Saturation,Mixture) color space. The position of these planes inside the RGB cube is modeled, based on the distribution of r, g and b cones of the human retina. To demonstrate the applicability of the proposed methodology, we present in this paper, the segmentation of "human skin" or "non-skin" pixels in digital color images. The performance of the Color Mixture was analyzed by a Gaussian distribution in the HSM, HSV and YCbCr color spaces. The method is compared with other skin/non-skin classifiers. The results demonstrate that our approach surpassed the performance of all compared methodologies. The main contributions of this paper are related to a new way for interpreting color of binary images, taking into account the bit-plane levels and the application in image processing techniques.

International Journal of Computers Communications & Control, 2016

This paper presents an unsupervised algorithm of colour image segmentation. This method combines the advantages of the approaches based on split and merge and region growing, and the use of the RGB and HSV colour representation models. The effectiveness of the proposed method has been verified by the implementation of the algorithm using three different testing images with homogeneous regions, spatially compact and continuous. It was observed that the proposed algorithm outperforms the other analysed techniques requiring shorter processing time when compared with the other methods.

Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, 2003

In this paper we describe a color image segmentation system that performs color clustering in a color space followed by a color region segmentation algorithm in the image domain. In color space, we describe two different algorithms that clusters similar colors using different measuring criteria and present our evaluation results on these two algorithms in comparison with three well-known color

In this paper we describe a color image segmentation system that performs color ,clustering in a color space followed by color region segmentation in the image domain. In color space, we describe two different algorithms that cluster similar colors using different criteria and present our evaluation results on these two algorithms in comparison with three well-known color segmentation algorithms. The region segmentation algorithm merges clusters in the image domain based on color similarity and spatial adjacency. We developed three different methods for merging regions in the image 'domain. The color image segmentation system has been implemented and tested on a variety of color images, including satellite images and moving car images. The system has shown to be both effective and efficient.