Rosalyn R. Porle

Preliminary study of Block Matching Algorithm (BMA) for video coding

The model of advance video coding technique can be divided into two main parts that are spatial model and temporal model. Spatial model exploit the redundancy in its single video frame (I frame) while temporal model exploit the redundancy among frames (P frame). Temporal model deals with motion estimation (ME) and motion compensation (MC) algorithm with the matching technique called “Block Matching Algorithm” (BMA) to produce the next encoded video frame with motion vector. In this paper, seven types of famous BMA technique such as Exhaustive Search, Three Step Search, New Three Step Search, Simple and Efficient Three Step Search, Four Step Search, Diamond Search and Adaptive Rood Pattern Search have been used to analyze the video frames quality with different block size and different sequence of I and P frame.

Comparison study of Hidden Markov Model gesture recognition using fixed state and variable state

This paper presents a method of gesture recognition using Hidden Markov Model (HMM). Gesture itself is based on the movement of each right hand (RH) and left hand (LH), which represents the word intended by the signer. The feature vector selected, gesture path, hand distance and hand orientations are obtained from RH and LH then trained using HMM to produce the respective gesture class. While training, in handling HMM state, we introduce fixed state and variable state, where in fixed state, the numbers of state is generally fixed for all gestures and while the number of state in variable state is determined by the movement of the gesture. It was found that fixed state gave the highest rate of recognition achieving 83.1%.

Fuzzy-rule-based object identification methodology for NAVI system

We present an object identification methodology applied in a navigation assistance for visually impaired (NAVI) system. The NAVI has a single board processing system (SBPS), a digital video camera mounted headgear, and a pair of stereo earphones. The captured image from the camera is processed by the SBPS to generate a specially structured stereo sound suitable for vision impaired people in understanding the presence of objects/obstacles in front of them. The image processing stage is designed to identify the objects in the captured image. Edge detection and edge-linking procedures are applied in the processing of image. A concept of object preference is included in the image processing scheme and this concept is realized using a fuzzy-rule base. The blind users are trained with the stereo sound produced by NAVI for achieving a collision-free autonomous navigation.

BMA Performance of Video Coding for Motion Estimation

The major advantages of Block Matching Algorithm is it can reduce the temporal redundancy by placing a direction and magnitude of a Motion Vector (MV) instead of an image block. There are various type of BMA for motion estimation technique have been proposed to enhance the accuracy of the MV of the image block and also to reduce the computational complexity of the BMA for time consuming. This paper investigate and presents a very brief introduction to seven type of famous BMA that are Exhaustive Search, Three Step Search, New Three Step Search, Simple and Efficient Three Step Search, Four Step Search, Diamond Search and Adaptive Rood Pattern Search. This paper also performs all of the seven BMA techniques with different parameter of I and P mode sequence, different search range and also different block size.

Wavelet-based skin segmentation for detecting occluded arms in human body pose modelling system

The task of correctly tracking the body parts is one of the crucial problems in the human body pose modelling. Various factors need to be investigated as the variety of body poses is unlimited and the visual appearance varies according to the environment. The human body can be composed into several parts such as the head, torso, arms and legs. The arms can be considered as the most challenging body part to be tracked since it tends to move fast and usually occluded within other body parts. This paper addresses the problem of extracting the arms which are occluded in the torso part. A wavelet-based skin segmentation method is applied to detect the skin region. The segmentation procedure is performed using six different colour spaces namely the RGB, rgb, HSI, TSL, SCT and CIELAB. The segmentation performances are evaluated on colour component basis. The aim of this paper is to determine the best colour components that are suitable for this segmentation procedure.

Multiple features integration in 2D upper human body pose modelling system

Human body pose modelling system is directly influenced by the image features used in the system, its model representation and also its application. This paper presents silhouette, edge and colour extraction methodology for detecting the human body parts in image sequences. Silhouette is used as input for head and torso pose estimation. Meanwhile, edge and colour are used respectively as input for non-occluded and occluded arms pose estimation. The body parts are estimated using histogram analysis and template matching techniques, which is then represented by rectangular shape.

A LabVIEW Design for Frontal and Non-Frontal Human Face Detection System in Complex Background

Biometric identification has advanced vastly since many decades ago. It became a blooming area for research as biometric technology has been used extensively in areas like robotics, surveillance, security and others. Face technology is more preferable due to its reliability and accuracy. By and large, face detection is the first processing stage that is performed before extending to face identification or tracking. The main challenge in face detection is the sensitiveness of the detection to pose, illumination, background and orientation. Thus, it is crucial to design a face detection system that can accommodate those problems. In this paper, a face detection algorithm is developed and designed in LabVIEW that is flexible to adapt changes in background and different face angle. Skin color detection method blending with edge and circle detection is used to improve the accuracy of face detected. The overall system designed in LabVIEW was tested in real time and it achieves accuracy about 97%.

Two dimensional human head and torso pose modelling using windowing techniques

Two dimensional human body pose modelling system detects the human body parts, estimates their posture and then models them in an image plane using specified shape. In this paper, two windowing techniques are presented and then compared for the human head and torso pose modelling. The first technique, namely Windowing technique I estimates the torso followed by the head of the human. In this technique, the size of the head and torso are manually computed and then the position of the targeted parts is determined. The second technique, namely Windowing technique II, estimates the head followed by the torso of the human. The size and the position of the targeted parts are determined automatically with the implementation of distant transform and several assumptions on human body size and position. The windowing techniques only requires silhouette image as input image. In experimentation, the size and the position of each body part are evaluated from 100 images in indoor environment. From the overall results, the Windowing technique II performs better in terms of correct size and position estimation.

An Adaptive Learning Radial Basis Function Neural Network for Online Time Series Forecasting

Most of the neural network based forecaster operated in offline mode, in which the neural network is trained by using the same training data repeatedly. After the neural network reaches its optimized condition, the training process stop and the neural network is ready for real forecasting. Different from this, an online time series forecasting by using an adaptive learning Radial Basis Function neural network is presented in this paper. The parameters of the Radial Basis Function neural network are updated continuously with the latest data while conducting the desired forecasting. The adaptive learning was achieved using the Exponential Weighted Recursive Least Square and Adaptive Fuzzy C-Means Clustering algorithms. The results show that the online Radial Basis Function forecaster was able to produce reliable forecasting results up to several steps ahead with high accuracy to compare with the offline Radial Basis Function forecaster.

Mandarin syllables speech trainer based on F1 and F2 formant frequencies

Computer based speech training system is one of the applications in speech technology that aims to provide a facility to learn foreign language. Language teaching nowadays increasingly focuses on pronunciation skills since the teaching standard within language teaching has shifted towards more emphasis on the ability to communicate orally. In this project, a system was developed to train user on pronouncing Mandarin syllables, by observing the first two formant frequencies, F1 and F2. These two values are obtained from Linear Prediction Coefficients and are affected by jaw opening and position of tongue, respectively. The user is requested to adjust these two aspects in order to improve the pronunciation. The level of accuracy of the syllable pronounced by user was obtained by comparing the formant frequencies with the standard pronunciation, which was pronounced by three Mandarin lecturers. The system was tested on a group of students and the result shows that these students can pronounce the syllables correctly in less trial compared to another group of students who were trying to imitate the syllable by just listening to the sample pronunciation.