Walaa M. Abd-Elhafiez

On Cluster Validity Indexes in Fuzzy and Hard Clustering Algorithms for Image Segmentation

This paper addresses the issue of assessing the quality of the clusters found by fuzzy and hard clustering algorithms. In particular, it seeks an answer to the question on how well cluster validity indexes can automatically determine the appropriate number of clusters that represent the data. The paper surveys several key existing solutions for cluster validity in the domain of image segmentation. In addition, it suggests two new indexes. The first one is based on Akaikes information criterion (AIC). While AIC was devoted to other domains such as statistical estimation of model fitting, it is implemented here for the first time as a validation index. The second index is developed from the well-established idea of cross-validation. The existing and new indexes are evaluated and compared on several synthetic images corrupted with noise of varying levels and volumetric MR data.

Object-based hybrid image and video coding scheme

Image and video data compression has been found to be necessary to reduce the amount of data represents to meet a bit rate requirement of the up to date communication networks. This paper describes two efficient image and video coding schemes. On one hand an efficient object-based hybrid image coding (OB-HIC) scheme, and on the other hand an object-based hybrid video coding (OB-HVC) scheme. The proposed schemes are based on using the discrete wavelet transform (DWT and 3-D DWT) in conjunction with the discrete cosine transform (DCT and 3-D DCT); (DWT and DCT with image, 3-D DWT and 3-D DCT with video), to provide coding performance superior to the popular image and video coders. The proposed methods use combination of the object-based DCT and 3-D DCT coding and the high performance of the set partitioning in hierarchical tree (SPIHT/3-D SPIHT) coding. The modification of the subband image data in the wavelet domain is done based on the DCT transformation and the classification of the coefficient in the low-frequency image and video subband (LL). The modification processes provide new subband image and video data containing almost the same information of the original one but having smaller values of the wavelet coefficient. Simulation results of the proposed methods demonstrate that, with small addition in the computational complexity of the coding process, the PSNR performance of the proposed algorithms are much higher than that of the SPIHT and 3-D SPIHT test coder and some of the famous image and video coding techniques.

Hybrid scheme for lifting based image coding

This paper describes an efficient lifting based hybrid image coding (LB-HIC) scheme. The proposed scheme is based on using the lifting transform (LT) in conjunction with the discrete cosine transform (DCT) to provide coding performance superior to the popular image coders. The proposed method uses combination of the DCT coding and the high performance of the set partitioning in hierarchical tree (SPIHT) coding. The subband image data in the lifting domain is modified based on the DCT and the object classification of the coefficient in the low-frequency image subband (LL). The modification process provides a new subband image data containing almost the same information of the original one but having smaller values of the lifting coefficients. Simulation results of the proposed method demonstrate that, with small addition in the computational complexity of the coding process, the peak signal-to-noise ratio (PSNR) performance of the proposed algorithm is much higher than that of the SPIHT test coder and some of famous image coding techniques.

Fast and accurate approaches for image and moving object segmentation

This paper introduces new two approaches for object extraction. The first one extracts the object in the image by separating the objects from the background. The second one extracts moving object in video sequences from the background. The approaches are based on background subtraction and edge information in the image. An initial spatial segmentation of the image is obtained by rainfalling watershed scheme. The resulting mask is used with a post-processing operation to remove the small regions and filling the holes of object region. After estimating the boundary of the object, the background subtraction process will be applied to get the final object. The proposed algorithm provides accurate results with different kinds of image. Simulation results demonstrated the efficiency of this method. In order to achieve object-based functionalities in image sequences, a new segmentation algorithm is applied to extract the video object planes (VOPs). The video approach is based on incorporating the spatial segmentation represented by fuzzy edge segmentation algorithm and temporal information represented by change detection. The simulation results demonstrated that the proposed algorithm can successfully extract moving objects from various sequences with different kinds of motion object. Nevertheless, the boundaries of the extracted object are not accurate enough to place them in different scenes.

Face identification system in video

This paper provides a simple and efficient method to detect human faces in videos and recognize persons within the video according to a preset database of known persons. The proposed system consists of three main steps. The first step is skin-like regions detection in CIE-Luv color space. The second step is face detection based on skin-like regions, contour detection and geometrical properties such as face shape. The third step is face verification, in which, each face is compared with a gallery of known faces and the location of the best matched one is returned. Face verification step uses variance formula and skin to non-skin percentage in each facial feature to compare the test face and the faces images in the known database. The proposed system was tested on many different videos with different number of persons in the video. The detected faces are compared with a preset database of known images. Experimental results show that the proposed system is efficient enough to detect faces in different lighting conditions, head pose and face expressions. The results of verification step show that the proposed system is able to retrieve faces from a database with good accuracy in a reasonable computational time compared with classical method, variance estimation method and facial extraction method.

An efficient scheme for face detection based on contours and feature skin recognition

Human face detection is very useful in many applications such as communications, automatic access control systems, video browsing, security control, verification of credit cards, identifying criminals and so on. This work provides a simple and efficient technique to detect human faces in still images. The new method based on skin color, contour drawing and feature extraction. The features under consideration are eyes, nose and mouth. The technique used to extract facial features developed based on feature location with respect to face dimensions. The proposed algorithm was tested on various images and its performance was found to be good in most cases. Experimental results show that our method of human face detection achieves very encouraging results with good accuracy, great speed and simple computations.

Effectiveness of the Image/Video Compression and Its Role in Supporting e-Learning Systems

In this paper, we address the problem of delivery of image and multimedia services on bandwidth deficient wire and wireless networks for E-learning. Became exponential growth in the volume of image and multimedia limits the ability of companies to manage this data, image and multimedia and control effectively, and with continued high storage costs make the companies having problems retrieving image and multimedia. This research aims to address these items discussion and analysis: concepts of image/video compression, the characteristics of image/video compression, the benefits and constraints of the use of image/video compression technique for E-learning lecture videos. Also, a efficient image and video compression technique is presented as a tool, which may be used in E-learning systems.

Non-embedded image coding techniques based on pixels classification

In this paper, new techniques for the DCT image coding based in pixels classifications are proposed. Two image coding approaches based on the object extraction are presented to study the effect of the object based image coding on the compression quality. Moreover, modification of the traditional JPEG method based on region-of-interest coding is achieved. In the beginning of all approaches, the image is subdivided into a block of pixels with block size of NxN. Firstly; the block must be classified as foreground or background block based on a preprocessing step. The foreground blocks will be compressed via JPEG technique but with significant quantized coefficients and the DC coefficient only from one block in the background is used to code it. The simulation result shows that the proposed technique provides competitive compression performance relative to the traditional JPEG technique.