Puteri Suhaiza Sulaiman

Skin Segmentation Using YUV and RGB Color Spaces

Skin detection is used in many applications, such as face recognition, hand tracking, and human-computer interaction. There are many skin color detection algorithms that are used to extract human skin color regions that are based on the thresholding technique since it is simple and fast for computation. The efficiency of each color space depends on its robustness to the change in lighting and the ability to distinguish skin color pixels in images that have a complex background. For more accurate skin detection, we are proposing a new threshold based on RGB and YUV color spaces. The proposed approach starts by converting the RGB color space to the YUV color model. Then it separates the Y channel, which represents the intensity of the color model from the U and V channels to eliminate the effects of luminance. After that the threshold values are selected based on the testing of the boundary of skin colors with the help of the color histogram. Finally, the threshold was applied to the input image to extract skin parts. The detected skin regions were quantitatively compared to the actual skin parts in the input images to measure the accuracy and to compare the results of our threshold to the results of other’s thresholds to prove the efficiency of our approach. The results of the experiment show that the proposed threshold is more robust in terms of dealing with the complex background and light conditions than others.

A FAST AND ACCURATE METHOD FOR AUTOMATIC CORONARY ARTERIAL TREE EXTRACTION IN ANGIOGRAMS

Coronary arterial tree extraction in angiograms is an essential component of each cardiac image processing system. Once physicians decide to check up coronary arteries from x-ray angiograms, extraction must be done precisely, fast, automatically and including whole arterial tree to help diagnosis or treatment during the cardiac surgical operation. This application is very helpful for the surgeon on deciding the target vessels prior to coronary artery bypass graft surgery. Some techniques and algorithms are proposed for extracting coronary arteries in angiograms. However, most of them suffer from some disadvantages such as time complexity, low accuracy, extracting only parts of main arteries instead of the full coronary arterial tree, need manual segmentation, appearance of artifacts and so forth. This study presents a new method for extracting whole coronary arterial tree in angiography images using Starlet wavelet transform. To this end, firstly we remove noise from raw angiograms and then sharpen the coronary arteries. Then coronary arterial tree is extracted by applying a modified Starlet wavelet transform and afterwards the residual noises and artifacts are cleaned. For evaluation, we measure proposed method performance on our created data set from 4932 Left Coronary Artery (LCA) and Right Coronary Artery (RCA) angiograms and compared with some state-of-the-art approaches. The proposed method shows much higher accuracy 96% for LCA and 97% for RCA, higher sensitivity 86% for LCA and 89% for RCA, higher specificity 98% for LCA and 99% for RCA and also higher precision 87% for LCA and 93% for RCA angiograms.

Echocardiography Image Segmentation: A Survey

Segmentation is an important step in medical imaging to acquire qualitative measurements such as the location of the desired objects and also for quantitative measurements such as area, volume or the analysis of dynamic behaviour of anatomical structures over time. Among these images, ultrasound images play a crucial role, because they can be produced on video-rate and therefore allows a dynamic analysis of moving structures. In addition, the acquisition of these images is non-invasive, cheap, and does not require ionizing radiations compared to other medical imaging techniques. On the other hand, the automatic segmentation of anatomical structures in ultrasound imagery is a real challenge due to acoustic interferences (speckle noise) and artifacts which are inherent in these images. This paper surveys the literature often recent researches on echocardiography image segmentation methods, focusing on techniques developed for medical. First, we present a review of articles by clinical application to highlight the approaches that have been investigated and degree of validation that has been done in different clinical domains. Then, we present a classification of methodology in terms of use of prior information. We conclude by selecting ten recent papers which have presented original ideas that have demonstrated particular clinical usefulness or potential specific to the echocardiography segmentation problem. The contribution of the paper is in three ways: 1) to serve as a tutorial on the field for both clinicians and technologists, 2) to provide an extensive account of segmentation techniques in a comprehensive and systematic manner, and 3) to critically review recent approaches in terms of their performance and degree of clinical evaluation with respect to the final goal of cardiac functional analysis.

Coronary Artery Segmentation in Angiograms with Pattern Recognition Techniques -- A Survey

Medical image processing is nowadays one of the best tools to make an informative model from a raw image of each part of the body, and segmentation is the most important step in which used to extract significant features. Coronary artery segmentation algorithm in angiograms is a fundamental component of each cardiac image processing system. There are lots of techniques and algorithms proposed for extracting coronary arteries in angiograms. But based on our knowledge, there is not any review paper to categorize and compare them together. In this paper, we have divided these algorithms into five major classes and propose a survey for the main class, pattern recognition, which is a famous technique in this manner. We studied all the papers in the pattern recognition class and defined six categories for them: (1) Multi scale approaches (2) Region growing approaches (3) Matching filters approaches (4) Mathematical morphology approaches (5) Skeleton based approaches and (6) Ridge based approaches. Finally, we made a table to compare all the algorithms in each category against criteria such as: user interaction, angiography types, dimensionality, enhancement method, full coronary artery output, whole tree output, and 3D reconstruction ability.

Hybrid Pixel-Based Method for Cardiac Ultrasound Fusion Based on Integration of PCA and DWT

Medical image fusion is the procedure of combining several images from one or multiple imaging modalities. In spite of numerous attempts in direction of automation ventricle segmentation and tracking in echocardiography, due to low quality images with missing anatomical details or speckle noises and restricted field of view, this problem is a challenging task. This paper presents a fusion method which particularly intends to increase the segment-ability of echocardiography features such as endocardial and improving the image contrast. In addition, it tries to expand the field of view, decreasing impact of noise and artifacts and enhancing the signal to noise ratio of the echo images. The proposed algorithm weights the image information regarding an integration feature between all the overlapping images, by using a combination of principal component analysis and discrete wavelet transform. For evaluation, a comparison has been done between results of some well-known techniques and the proposed method. Also, different metrics are implemented to evaluate the performance of proposed algorithm. It has been concluded that the presented pixel-based method based on the integration of PCA and DWT has the best result for the segment-ability of cardiac ultrasound images and better performance in all metrics.

Removing Shadow for Hand Segmentation Based on Background Subtraction

Hand segmentation is an important stage for accurate hand detection and background subtraction is one of the best solutions to detect the hand motion accurately, however the shadow is the critical problem in this technique which is not easy to separate the hand region from the shadow area. Removing shadow using an automatic threshold will be a good solution to detect the hand region where the variety of skin color and lighting condition affect the hand segmentation. The proposed approach involves three stages: First, we convert RGB color model to YUV space to get the benefit of separation the luminance channel (Y) from the chrominance channels (U, V) to reduce the effect of shadow, reflections and, etc. In the second stage, we applied background subtraction technique to the V channel to remove the unwanted background noise and to get the hand and shadow pixels. Finally, we used shareholding technique by considering a mean value of the pixels of foreground image (the hand and shadow pixels) as automatic threshold value and other tow static thresholds to distinguish the hand region from shadow pixels. After background subtraction, we used the famous morphology techniques (Erosion and Dilation) to enhance the accuracy of hand detection. We measure the accuracy for the results by compare the detect hand pixels to the actual hand pixels quantitatively. From the results, we noticed that our proposed approach is accurate and suitable for real time application systems.

Low Cost Skin Segmentation Scheme in Videos Using Two Alternative Methods for Dynamic Hand Gesture Detection Method

Recent years have witnessed renewed interest in developing skin segmentation approaches. Skin feature segmentation has been widely employed in different aspects of computer vision applications including face detection and hand gestures recognition systems. This is mostly due to the attractive characteristics of skin colour and its effectiveness to object segmentation. On the contrary, there are certain challenges in using human skin colour as a feature to segment dynamic hand gesture, due to various illumination conditions, complicated environment, and computation time or real-time method. These challenges have led to the insufficiency of many of the skin color segmentation approaches. Therefore, to produce simple, effective, and cost efficient skin segmentation, this paper has proposed a skin segmentation scheme. This scheme includes two procedures for calculating generic threshold ranges in Cb-Cr colour space. The first procedure uses threshold values trained online from nose pixels of the face region. Meanwhile, the second procedure known as the offline training procedure uses thresholds trained out of skin samples and weighted equation. The experimental results showed that the proposed scheme achieved good performance in terms of efficiency and computation time.

Age Range Estimation Based on Facial Wrinkle Analysis Using Hessian Based Filter

Aging is a normal process that has an effect on different parts of the human body under the influence of various biological and environmental aspects. The most prominent changes that occur on the face are the form of the skin wrinkles, which are the main objective of this research. Specific wrinkle detection is an important task in face textural analysis. Previously, some researchers have been proposed the age range estimation based on wrinkle analysis in literature, but poor localization limits the performance of the whole age estimation process. This is because, when less number of wrinkles are detected or extracted, it will consequently affect the process to estimate the correct age. Therefore, we address this issue to enhance age range estimation method using a new approach to extract correct facial wrinkles for further analysis. We propose a method to extract facial wrinkle in face image using Hessian based filter (HBF) for age estimation. In other word, this research focus on age range estimation method based on facial wrinkle analysis extracted from facial image obtained from FG-NET database using hessian based filter. The proposed filter is theoretically straightforward, however, it significantly increases the wrinkle analysis result compared to previous methods. The result shows that HBF successfully obtained higher accuracy with over 90 % estimation rate.

Filling Sharp Features on Corner of Triangular Mesh by using Enhanced Advancing Front Mesh (EAFM) method

Repairing an incomplete polygon mesh constitutes a primary difficulty in 3D model construction, especially in the computer graphics area. The objective of hole-filling methods is to keep surfaces smoothly and continually filled at hole boundaries while conforming with the shapes. The Advancing Front Mesh (AFM) method was normally used to fill simple holes. However, there has not been much implementation of AFM in handling sharp features. In this paper, we use an AFM method to fill a holes on sharp features. The Enhanced Advancing Front Mesh (EAFM) method was introduced when there was a conflict during triangle creation. The results of the study show that the presented method can effectively improve the AFM method, while preserving the geometric features and details of the original mesh.

Tricuspid Valve Extraction in Transesophageal Echocardiography

In the past decades, echocardiography has appeared as an important modality in medical field to assess heart’s function and structures as well as for diagnosis and evaluation. Many image processing researches are done to enhance the imaging aspect and produce better quality of image. Numerous research have been conducted on mitral valve, but only a few on the geometry or annular dynamics of the tricuspid valve. Accurateness in measuring and reconstructing tricuspid valve is an important issue, not only for surgical decision-making process but also in deciding the suitable surgical technique on patient such as valve implication or ring placement. In this paper, we will discuss on techniques that have been applied recently in measuring and modelling tricuspid valve and as for experiment, 3DTEE image was used using level set technique discussed in this paper. Our findings will be focusing more on those techniques applied on 3D echocardiography images from different angels and positions.