Please use this identifier to cite or link to this item:
Title: Multi-Objective Optimization Based Image Segmentation: Method and Applications
Keywords: Image segmentation, Histogram, Multi-objective optimization, Crystal Size Distribution (CSD), Breast ultrasound
Issue Date: 19-Jan-2012
Citation: PERIASAMY KARTHIK RAJA (2012-01-19). Multi-Objective Optimization Based Image Segmentation: Method and Applications. ScholarBank@NUS Repository.
Abstract: Image analysis plays a crucial role in various fields such as biology, medicine, remote sensing, robotics and manufacturing. Image segmentation is a critical step in image analysis since the result of segmentation plays an important role in feature extraction. In this work, image segmentation is carried out by thresholding. Generally, the threshold is selected by optimizing a single objective. Thresholding can be improved by combining the objectives of two different methods (Otsu and minimum error thresholding methods). Hence, in this work, the optimum threshold is calculated by solving a multi-objective optimization (MOO) problem. The two objectives used in this work are maximizing the between-class variance and the minimizing the error while histogram fitting. This MOO is solved using the plain aggregating approach and simulated annealing by assigning appropriate weights to each objective function. The MOO based thresholding overcomes the limitations of the individual approaches and outperforms the results obtained by thresholding using either of the single objectives. The misclassification rate of the MOO approach is compared with the traditional Otsu and minimum error thresholding methods. The MOO based approach is tested on several examples. The first application is in the estimation of crystal size distribution (CSD) using Particle Vision and Measurement (PVM) images to assist in crystallization process control. In this study, the segmentation results of the developed method are compared with the results of Otsu and minimum error method. The segmented images are further processed by means of feature extraction to estimate the CSD. The algorithm is tested on a set of artificially generated crystallization images. The accuracy of this algorithm is gauged by comparing the CSD estimated to the data used to generate the artificial images. This accuracy was found to be around 92% for images in which about 20 - 25 particles exist. The effect of parameters such as the number of images, the number of particles in the images, and noise level in the images on the estimated CSD is investigated. The second application relates to classifying benign and malignant tumors to assist radiologists involved in the treatment of cancer patients. Our proposed MOO methodology is used to segment the tumors (regions of interest) and the results are compared with the other methods. With the help of feature extraction, a set of required features are extracted from the images. These features can then be used by radiologists for classification purposes and subsequent treatment. In addition to the two abovementioned process and medical applications, other illustrative examples are also included to illuminate the utility of the proposed MOO based thresholding in aiding decision making for real-world applications.
Appears in Collections:Master's Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
PeriasamyKR.pdf1.9 MBAdobe PDF



Page view(s)

checked on Nov 3, 2018


checked on Nov 3, 2018

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.