P. K. Nanda

Unsupervised Color Image Segmentation Using Compound Markov Random Field Model

In this paper, we propose an unsupervised color image segmentation scheme using homotopy continuation method and Compound Markov Random Field (CMRF) model. The proposed scheme is recursive in nature where model parameter estimation and the image label estimation are alternated. Ohta (I 1, I 2, I 3) model is used as the color model for image segmentation and we propose a compound MRF model taking care of intra-color and inter-color plane interactions. The CMRF model parameters are estimated using Maximum Conditional Pseudo Likelihood (MCPL) criterion and the MCPL estimates are obtained using homotopy continuation method. The image label estimation is formulated using Maximum a Posteriori criterion and the MAP estimates are obtained using hybrid algorithm. In the context of misclassification error, the proposed unsupervised scheme with CMRF model exhibited improved segmentation accuracy as compared to MRF model and Katos method.

Constrained Markov Random Field Model for Color and Texture Image Segmentation

In this paper, the problem of color image segmentation is addressed as a pixel labeling problem. The observed color image is assumed to be the degraded version of the image labels. We have proposed a new Markov random field (MRF) model known as constrained MRF (CMRF) model to model the unknown image labels and Ohta (I1I2I3) model is used as the color model. The unique feature of the proposed CMRF model is found to posses a unifying feature of modeling scene and texture images as well. The labels are estimated using maximum a posteriori (MAP) estimation criterion. A hybrid algorithm is proposed to obtain the MAP estimate and the performance of the algorithm is found to be better than that of using simulated annealing (SA) algorithm. The performance of the proposed model is compared with JSEG method and the proposed model is found to be better than JSEG method.

Constrained Compound Markov Random Field Model with Graduated Penalty Function for Color Image Segmentation

In this paper, an unsupervised color image segmentation scheme has been proposed for preserving strong and weak edges as well. A Constrained Compound Markov Random Field (MRF) has been proposed as the a priori model for the color labels. We have used Ohta (I1, I2, I3) color model and a controlled correlation of the color space has been accomplished by the proposed compound MRF model. The Constrained Compound MRF (CCMRF) is found to possess the unifying property of modeling scenes as well as color textures. In unsupervised scheme, the associated model parameters and the image labels are estimated recursively. The model parameters are the Maximum Conditional Pseudo Likelihood (MCPL) estimates and the labels are the Maximum a Posteriori (MAP) estimates. The performance of the proposed scheme has been compared with that of Yu’s method and has been found to exhibit improved performance in the context of misclassification error.

Constrained compound Markov random Field Model for segmentation of color texture and scene images

In this paper, we propose a constrained compound Markov random field model (MRF) to model color texture as well as scene images. Ohta (I1, I2, I3) color model is used as the color model for segmentation. Besides, intra plane model, the constrained model is modified to take care of inter-plane interaction as well. Hence, the model is called as double constrained compound MRF (DCCMRF) model. The problem is formulated as pixel labelling problem and the pixel labels are estimated using maximum a posteriori (MAP) criterion.The MAP estimates are obtained using hybrid algorithm. The DCCMRF model exhibited improved segmentation accuracy as compared to DCMRF, MRF, double MRF (DMRF), double Gauss MRF(DGMRF) and JSEG method. The proposed models have been successfully tested for two, four and five class problem.

Unsupervised Color Image Segmentation Using Constrained Compound MRF Model with Bi-level Line Field

In this paper, we propose an unsupervised color image segmentation scheme using homotopy continuation method and Compound Markov Random Field (CMRF) model with Bilevel Binary Line Fields. The scheme is specifically meant to preserve weak edges besides the well defined strong edges. The proposed scheme is recursive in nature where model parameter estimation and the image label estimation are alternated. Ohta (I1, I2, I3) model is used as the color model for image segmentation and we propose a compound MRF model taking care of intra-color and inter-color plane interactions. The CMRF model parameters are estimated using Maximum Conditional Pseudo Likelihood (MCPL) criterion and the MCPL estimates are obtained using homotopy continuation method. The image label estimation is formulated using Maximum a Posteriori criterion and the MAP estimates are obtained using hybrid algorithm. In the context of misclassification error, the proposed unsupervised scheme with CMRF model exhibited improved segmentation accuracy as compared to Yu and Clausi ’s method.

Color image segmentation using constrained compound Markov Random Field model and homotopy continuation method

In this paper, we propose a supervised color image segmentation using homotopy continuation method and Markov random field (MRF) model. We propose a constrained compound MRF model to take care of color texture and scene images. Ohta (I1, I2, I3) model is used as the color for image segmentation. We also have extended the proposed model to inter-color-planes as well as intra-color-planes of the color model and thus a double constrained compound MRF (DCCMRF) model is proposed. The a priori MRF model parameters are estimated using the proposed homotopy continuation based method. The model parameters are the maximum pseudo likelihood estimates. The DCCRMRF model with estimated model parameters exhibited improved segmentation accuracy as compared to DCMRF, MRF, double MRF (DMRF) and JSEG method.

Constrained Compound MRF Model with Bi-Level Line Field for Color Image Segmentation

Image segmentation is a basic early vision problem which serves as precursor to many high level vision problems. Color image segmentation provides more information while solving high level vision problems such as, object recognition, shape analysis etc. Therefore, the problem of color image segmentation has been addressed more vigorously for more than one decade. Different color models such as RGB, HSV, YIQ, Ohta (I1, I2, I3), CIE(XYZ, Luv, Lab) are used to represent different colors [5]. From the reported study, HSV and (I1, I2, I3) have been extensively used for color image segmentation. Ohta color space is a very good approximation of the Karhunen-Loeve transformation of the RGB, and is very suitable for many image processing applications [1]. Image Modeling plays a crucial role in image anal‐ ysis. Stochastic models, particularly MRF models, have been successfully used as the image model for image restoration and segmentation [2], [3], [4]. MRF model has also been success‐ fully used as the image model while addressing the problem of color image segmentation both in supervised and unsupervised framework. Kato et al [6] have proposed a MRF model based unsupervised scheme for color image segmentation. In Kato s method, the model pa‐ rameters have been estimated using Maximum Likelihood criterion and the only parameter identified by the user is the number of class. This algorithm could be validated using differ‐ ent color textures and real images. Another color texture unsupervised segmentation algo‐ rithm has been proposed by Deng et al [7] and the method has been retermed as JSEG method. Recently, an unsupervised image segmentation algorithm has been proposed by Guo et al [8] where K-means has been used to initialize the classification in the classification of numbers. Very recently Scarpa et al. [13] have proposed a multiscale texture model and a related algorithm for the unsupervised segmentation of color images. In this scheme, the feature vectors have been collected and based on the feature vector the textures are then re‐

Supervised color image segmentation using constrained compound MRF model with Bi-level line Field

In this paper, we propose an supervised color image segmentation scheme using homotopy continuation method and Compound Markov Random Field (CMRF) model with Bilevel Binary Line Fields. The scheme is specifically meant to preserve weak edges besides the well defined strong edges. Ohta (I 1 , I 2 , I 3 ) model is used as the color model for image segmentation and we propose a compound MRF model taking care of intra-color and inter-color plane interactions. The CMRF model parameters are estimated using Maximum Conditional Pseudo Likelihood (MCPL) criterion and the MCPL estimates are obtained using homotopy continuation method. The image label estimation is formulated using Maximum a Posteriori criterion and the MAP estimates are obtained using hybrid algorithm. In the context of misclassification error, the proposed unsupervised scheme with CMRF model exhibited improved segmentation accuracy as compared to Yu and Clausi s method.