Showing papers by "Hamid R. Tizhoosh published in 2003"

Using reinforcement learning for image thresholding

Abstract: This paper introduces a reinforcement-learning concept to find the optimal threshold for digital images. The proposed approach can integrate human expert knowledge in an objective or subjective way to overcome the shortcomings of existing methods.

Filter fusion for image enhancement using reinforcement learning

Abstract: A new approach to image enhancement based on fusion of a number of filters using a reinforcement learning scheme is presented. In most applications the result of applying a single filter is usually unsatisfactory. Appropriate fusion of the results of several different filters, such as median, local average, sharpening, and Wiener filters, can resolve this difficulty. Many different techniques already exist in literatures. In this work, a reinforcement-learning agent will be proposed. During learning, the agent takes some actions (i.e., different weights for filters) to change its environment (the image quality). Reinforcement is provided by a scalar evaluation determined subjectively by the user. The approach has several advantages. The user interaction eliminates the need for objective image quality measures. No formal user model is required. Finally, no training data is necessary. The paper describes the implementation and evaluation of a global reinforced adjustment of the weights of the different filters.

Prostate's boundary detection in transrectal ultrasound images using scanning technique

Abstract: Prostate cancer is one of the leading causes of cancer death in men. Early detection of prostate cancer is very essential for the success of the treatment. Ultrasound B-mode images is the standard mean for imaging the prostate. The manual analysis of the ultrasound images consumes much time and effort. It is necessary to develop an automated algorithm to analyze the ultrasound images. The first important step in detecting the cancer is to detect the boundary of the prostate itself and to extract it from the image for further analysis. In this paper a multistage algorithm for prostate boundary detection is proposed. In the first stage, the proposed algorithm starts with enhancing the contrast of the image by sticks technique followed by smoothing the image by gauss kernel. In the second stage, scanning the image and applying knowledge base rules to find a seed point inside the prostate is implemented. This seed point is used to remove the false edges. Then by using a morphological opening algorithm, the remaining false edges can be removed. The final step is to use the seed point to scan the image in radial directions to find the prostate' s boundary.

Fast and Robust Fuzzy Edge Detection

Abstract: In recent years, fuzzy techniques have been applied to develop new edge detection techniques because they offer a flexible framework for edge extraction with respect to specific requirements. These techniques, however, are usually expensive in computing compared to classical approaches like the Sobel operator. In many practical applications we need fast edge detection. In this chapter, several fast methods are proposed which are suitable for cases where a rough edge estimation is required. On the other side, the result of edge detection techniques in noisy environments is often not satisfactory. In this chapter, also a robust algorithm based on fuzzy if-then rules is proposed that can detect edges and lines in noisy images.

Choquet integral-based aggregation of image template matching algorithms

Abstract: Template matching algorithms determine the best matching position of a reference image (template) on a larger image (scene) in either complete or incomplete information environment. In this work, our main objective is to devise a fuzzy integral-based aggregation scheme in an attempt to get more accurate and robust matching, by combining the matching decisions of a finite number of image template matching algorithms, Particularly, Choquet integrals associated with fuzzy measures can be used for handling fuzziness due to incomplete image information. In the present work, a fuzzy integral-based aggregated template matching system is developed on the basis of Choquet integral using belief, plausibility, and probability measure, while being interpreted as an optimistic, a pessimistic, and a noninteracting aggregation, respectively. Finally, to show a validation of Choquet integral-based template matching methods, three individual template matching methods (i,e., MOAD-matcher, SOAD-matcher, and SOSD-matcher) are combined using Choquet integral with respect to different fuzzy measures. Then, performance of these aggregated matchers is compared to individual matchers' performance. It is found that in a complementary sense a Choquet integral-based aggregation of template matching methods gives a better performance compared to the performance of the individual methods.