Wan Mimi Diyana Wan Zaki

Real-time background subtraction for video surveillance: From research to reality

This paper reviews and evaluates performance of few common background subtraction algorithms which are medianbased, Gaussian-based and Kernel density-based approaches. These algorithms are tested using four sets of image sequences contributed by Wallflower datasets. They are the image sequences of different challenging environments that may reflect the real scenario in video surveillances. The performances of these approaches are evaluated in terms of processing speed, memory usage as well as object segmentation accuracy. The results demonstrate that Gaussian-based approach is the best approach for real-time applications, compromising between accuracy and computational time. Besides, this paper may provide a better understanding of algorithm behaviours implemented in different situation for real-time video surveillance applications.

Design and development of a content-based medical image retrieval system for spine vertebrae irregularity

BackgroundContent-based medical image retrieval (CBMIR) system enables medical practitioners to perform fast diagnosis through quantitative assessment of the visual information of various modalities.MethodsIn this paper, a more robust CBMIR system that deals with both cervical and lumbar vertebrae irregularity is afforded. It comprises three main phases, namely modelling, indexing and retrieval of the vertebrae image. The main tasks in the modelling phase are to improve and enhance the visibility of the x-ray image for better segmentation results using active shape model (ASM). The segmented vertebral fractures are then characterized in the indexing phase using region-based fracture characterization (RB-FC) and contour-based fracture characterization (CB-FC). Upon a query, the characterized features are compared to the query image. Effectiveness of the retrieval phase is determined by its retrieval, thus, we propose an integration of the predictor model based cross validation neural network (PMCVNN) and similarity matching (SM) in this stage. The PMCVNN task is to identify the correct vertebral irregularity class through classification allowing the SM process to be more efficient. Retrieval performance between the proposed and the standard retrieval architectures are then compared using retrieval precision (Pr@M) and average group score (AGS) measures.ResultsExperimental results show that the new integrated retrieval architecture performs better than those of the standard CBMIR architecture with retrieval results of cervical (AGS > 87%) and lumbar (AGS > 82%) datasets.ConclusionsThe proposed CBMIR architecture shows encouraging results with high Pr@M accuracy. As a result, images from the same visualization class are returned for further used by the medical personnel.

Enhancement of background subtraction techniques using a second derivative in gradient direction filter

A new approach was proposed to improve traditional background subtraction (BGS) techniques by integrating a gradient-based edge detector called a second derivative in gradient direction (SDGD) filter with the BGS output. The four fundamental BGS techniques, namely, frame difference (FD), approximate median (AM), running average (RA), and running Gaussian average (RGA), showed imperfect foreground pixels generated specifically at the boundary. The pixel intensity was lesser than the preset threshold value, and the blob size was smaller. The SDGD filter was introduced to enhance edge detection upon the completion of each basic BGS technique as well as to complement the missing pixels. The results proved that fusing the SDGD filter with each elementary BGS increased segmentation performance and suited postrecording video applications. Evidently, the analysis using Fscore and average accuracy percentage proved this, and, as such, it can be concluded that this new hybrid BGS technique improved upon existing techniques.

Moving object detection using keypoints reference model

This article presents a new method for background subtraction (BGS) and object detection for a real-time video application using a combination of frame differencing and a scale-invariant feature detector. This method takes the benefits of background modelling and the invariant feature detector to improve the accuracy in various environments. The proposed method consists of three main modules, namely, modelling, matching and subtraction modules. The comparison study of the proposed method with a popular Gaussian mixture model proved that the improvement in correct classification can be increased up to 98% with a reduction of false negative and true positive rates. Beside that the proposed method has shown great potential to overcome the drawback of the traditional BGS in handling challenges like shadow effect and lighting fluctuation.

Multi-level segmentation method for serial computed tomography brain images

This paper presents an automated computed tomography brain segmentation approach used to segment intracranial into brain matters and cerebrospinal fluid in order to detect any asymmetry present. Intracranial midline is used as reference axial where left and right segmented regions are subjectively compared. Two-level Otsu multi-thresholding method has been developed and applied to 213 abnormal cases of serial computed tomography brain images of thirty one patients. Prior to that, multilevel Fuzzy C-Means is used to extract the intracranial from background and skull. The segmented regions found to be very useful in providing information regarding normal and abnormal structures in the intracranial where any asymmetry detected would indicate high probability of abnormalities. This approach proved to effectively isolate important homogenous regions of computed tomography brain images from which extracted features would provide a strong basis in the application of content-based medical image retrieval.

A New Approach of Skull Fracture Detection in CT Brain Images

This work demonstrates a new automated approach to segment skull from 2D-CT brain image to detect any fracture case. The key steps in the proposed approach include image normalization, centroid identification, multi-level global segmentation and skull skeletonization. Feature vectors such as location and fracture size are then extracted to represent fracture cases. Twenty eight encephalic fracture images are queried from a database of 3032 normal and fractured CT brain images to evaluate the usefulness of the skull segmentation as well as the extracted feature vectors in content-based medical image retrieval system (CBMIR). Retrieval performance of Normalized Euclidean and Normalized Manhattan distance metrics show almost perfect average recall-precision plots that portray the suitability of this approach to the CBMIR of fracture cases.

A review on the diabetic retinopathy assessment based on retinal vascular tortuosity

Diabetic Retinopathy (DR) is a severe vision threatening disease which causes visual loss for most of diabetic eye disease patients. This paper presents an overview of various methods used by other researchers in determining the retina vascular tortuosity and its association with DR. In this paper, several quantitative methods of retinal vascular tortuosity in Retinopathy of Premature (ROP) assessment are included as a benchmark of tortuosity measurement in DR. Initial finding suggests that there is a possible association between retinal vascular tortuosity and the development of DR. To confirm this, further research has to be conducted to see the frequency or rate of retinal vascular tortuosity. As such, this paper intends to report the preliminary work conducted which summarizes several developments in recent literature and discusses the various methods used for tortuosity measurement in DR.

Entrepreneurship Education in UKM: Essential Skills for First-Year Students

Most universities have developed entrepreneurial learning known as academic entrepreneurship (AE) by introducing entrepreneur-related courses for cross-discipline students. In order to increase student’s opportunities in developing business skills, Institute of Higher Education (IPT) in Malaysia encourages entrepreneurial program in higher academic institutions. The government supports the effort to reach the vision of Malaysia’s Economic Transformation Program (ETP) by 2020 in enhancing entrepreneurship abilities. Moreover, academic entrepreneurship has rooted in many global universities whereby the entrepreneurship subjects are introduced across the board regardless of faculties in order to provide basic entrepreneurial knowledge using adaptable learning pattern. The faculty entrepreneurship concept is applied in the teaching and learning process at Universiti Kebangsaan Malaysia (UKM) by providing entrepreneurial courses for first-year students from all faculties. The course comprises Fundamentals of Entrepreneurship and Innovation and was initiated 5 years ago involving 12 faculties in UKM. These papers introduce basic elements of entrepreneurship and related skills to students in different disciplines. Its main objective is to promote and generate business interests and ideas among students by providing fundamental knowledge of entrepreneurship so that a career in entrepreneurship is deemed as a feasible career option. Changing mind sets of students to be a job creator rather than a job seeker is today’s challenge that can be overcome by dissemination of invaluable business tools and materials which is the core aspect of this course. Concepts and theories of entrepreneurship including team building, teaming and leadership, strategy and management aptitudes, marketing and market research, financial and legal principles, manufacturing or production processes, and oral presentation skills will be taught. The link between various components of a business will be demonstrated and identified through business simulation games performed by students themselves. Series of periodic seminars and recorded videos displaying experiences of local successful entrepreneurs are presented to inspire students to embrace values and challenges of an entrepreneur. A business pitching and business concept poster competitions are held to assess students’ understanding and reflection of the course content translated into these forms of “hands-on” activities.

X-Ray image enhancement for anterior Osteophyte diagnosis

Computer assisted system has been implemented virtually in many domains within the field of medical imaging. An automated assessment system based on computer is important for early detection and diagnosis of various diseases. Moreover, image enhancement in medical imaging is a vital component in improving the quality of medical image, so that pre-screening of the disease can be carried accurately without performing any open surgery. Hence, pre-processing module is an important step to enhance the medical image, especially for X-ray modality, which normally affected by artefacts, noise and hardware limitations. A low quality of X-ray image will usually lead to inaccurate diagnosis. Therefore, a semi-automated assessment system has been developed to diagnose Anterior Osteophytes (AO) condition based on X-ray image that focused on cervical vertebrae. The system consists of four main modules: image enhancement, image segmentation, feature extraction and classification. This system is intended to facilitate medical practitioners in screening the AO condition with more accurate and faster diagnosis. A graphical user interface has also been developed to help the medical practitioners to perform the image enhancement easily. The proposed system is tested with 100 cervical X-ray images and the obtained accuracy is 60%. Finally, this system is suitable for real time implementation and can be implemented on most recent desktop technology.

On analyzing colour constancy approach for improving SURF detector performance

Robust key point detector plays a crucial role in obtaining a good tracking feature. The main challenge in outdoor tracking is the illumination change due to various reasons such as weather fluctuation and occlusion. This paper approaches the illumination change problem by transforming the input image through colour constancy algorithm before applying the SURF detector. Masked grey world approach is chosen because of its ability to perform well under local as well as global illumination change. Every image is transformed to imitate the canonical illuminant and Gaussian distribution is used to model the global change. The simulation results show that the average number of detected key points have increased by 69.92%. Moreover, the average of improved performance cases far out weight the degradation case where the former is improved by 215.23%. The approach is suitable for tracking implementation where sudden illumination occurs frequently and robust key point detection is needed.