Abd Rahman Ramli

Boosting diagnosis accuracy of Alzheimer's disease using high dimensional recognition of longitudinal brain atrophy patterns

OBJECTIVE Boosting accuracy in automatically discriminating patients with Alzheimers disease (AD) and normal controls (NC), based on multidimensional classification of longitudinal whole brain atrophy rates and their intermediate counterparts in analyzing magnetic resonance images (MRI). METHOD Longitudinal percentage of brain volume changes (PBVC) in two-year follow up and its intermediate counterparts in early 6-month and late 18-month are used as features in supervised and unsupervised classification procedures based on K-mean, fuzzy clustering method (FCM) and support vector machine (SVM). The most relevant features for classification are selected using discriminative analysis (DA) of features and their principal components (PC). Accuracy of the proposed method is evaluated in a group of 30 patients with AD (16 males, 14 females, age±standard-deviation (SD)=75±1.36 years) and 30 normal controls (15 males, 15 females, age±SD=77±0.88 years) using leave-one-out cross-validation. RESULTS Results indicate superiority of supervised machine learning techniques over unsupervised ones in diagnosing AD and withal, predominance of RBF kernel over lineal one. Accuracies of 83.3%, 83.3%, 90% and 91.7% are achieved in classification by K-mean, FCM, linear SVM and SVM with radial based function (RBF) respectively. CONCLUSION Evidence that SVM classification of longitudinal atrophy rates may results in high accuracy is given. Additionally, it is realized that use of intermediate atrophy rates and their principal components improves diagnostic accuracy.

Modeling of oil palm fruit maturity for the development of an outdoor vision system.

Color is the most important indicator farmers use to determine the maturity of the oil palm fruit called fresh fruit bunches (FFB) in the manual harvesting process. To automate the harvesting operation, the development of a vision system will replace the human eye for mature FFB recognition. In real plantation environments, variations in the daylight caused the light intensity to change, thus becoming the main issue that affects the automatic recognition process. In this study, the matured FFB was captured using a Sony digital Handycam on the day shift period. At the same time period of daylight intensity, a unit on foot candles (FC) also was simultaneously recorded using an Extech lightmeter data logger. From the linear regression analysis process, the mathematical model shows that there is a linear change between daylight intensity with the pixel value of the components green and blue. For the pixel value of the red component, the value will be linear at a maximum of 255 and at a certain intensity. To validate the mathematical model, this equation is used in the development of software for outdoor recognition processes.

MRI segmentation of Medical images using FCM with initialized class centers via genetic algorithm

Image segmentation is a critical stage in many computer vision and image process applications. Accurate segmentation of medical images is very essential in Medical applications but it is very difficult job due to noise and in homogeneity. Fuzzy C-Mean (FCM) is one of the most popular Medical image clustering methods. We noticed that for some images, FCM is sensitive to initialization of centre of clusters. This article introduced a new method based on the combination of genetic algorithm and FCM to solve this problem. The genetic algorithm is used to find initialized centre of the clusters. In this method, the centre is obtained by minimizing an object Function. This object Function specifies sum of distances between each data and their cluster centres. Then FCM is applied with to the case. The experimental result demonstrates the effectiveness of new method by able to initialize centre of the clusters.

Denoising of natural images through robust wavelet thresholding and genetic programming

Digital images play an essential role in analysis tasks that can be applied in various knowledge domains, including medicine, meteorology, geology, and biology. Such images can be degraded by noise during the process of acquisition, transmission, storage, or compression. The use of local filters in image restoration may generate artifacts when these filters are not well adapted to the image content as a result of the heuristic optimization of local filters. Denoising methods based on learning procedure are more capable than parametric filters for addressing the conflicts between noise suppression and artifact reduction. In this study, we present a nonlinear filtering method based on a two-step switching scheme to remove both salt-and-pepper and additive white Gaussian noises. In the switching scheme, two cascaded detectors are used to detect noise, and two corresponding estimators are employed to effectively and efficiently filter the noise in an image. In the process of training, a method according to patch clustering is utilized, and genetic programming (GP) is subsequently applied to determine the optimum filter (wavelet-domain filter) for each individual cluster, while in testing part, the optimum filter trained beforehand by GP is recovered and used on the inputted corrupted patch. This adaptive structure is employed to cope with several noise types. Experimental and comparative analysis results show that the denoising performance of the proposed method is superior to that of existing denoising methods as per both quantitative and qualitative assessments.

Just-in-time outdoor color discrimination using adaptive similarity-based classifier

The color recognition and identification in operation time is a critical task in color-based computer vision applications. The main problem for recognizing the real color arises when the color characteristics are changed dynamically in the life time of a system. The outdoor color models which have been addressed by some researchers have serious practical limitations to employ in real applications. Moreover, due to high fluctuations in environment illumination, using conventional classifier for discriminating colors is a complicated task. In this paper, a just-in-time and model-free solution in order to discriminate outdoor colors on data driven modality is proposed. For this purpose, adaptive similarity-based classifier is utilized to track the colors data evolution during a day.

Improvement of digital mammogram images using histogram equalization, histogram stretching and median filter

Breast cancer is one of the most important diseases in females worldwide. According to the Malaysian Oncological Society, about 4% of women who are 40 years old and above are involved have breast cancer. Masses and microcalcifications are two important signs of breast cancer diagnosis on mammography. Enhancement techniques, i.e. histogram equalization, histogram stretching and median filters, were used to provide better visualization for radiologists in order to help early detection of breast abnormalities. In this research 60 digital mammogram images which includes 20 normal and 40 confirmed diagnosed cancerous cases were selected and manipulated using the mentioned techniques. The original and manipulated images were scored by three expert radiologists. Results showed that the selected methods have a positive significant effect on image quality.

Natural image noise level estimation based on local statistics for blind noise reduction

This study proposes an automatic noise estimation method based on local statistics for additive white Gaussian noise. Noise estimation is an important process in digital imaging systems. For example, the performance of an image denoising algorithm can be significantly degraded because of poor noise level estimation. Most of the literature on the subject tends to use the true noise level of a noisy image when suppressing noise artifacts. Moreover, even with the given true noise level, these denoising techniques still cannot attain the best result, particularly for images with complicated details. In this study, a patch-based estimation technique is used to estimate for noise level and applies it to the proposed blind image denoising algorithm. Our approach includes selecting low-rank sub-image with removing high-frequency components from the contaminated image. This selection is according to the gradients of patches with the same statistics. Consequently, we need to estimate the noise level from the selected patches using principal component analysis (PCA). For blind denoising applications, the proposed denoising algorithm integrates the undecimated wavelet-based denoising algorithms and PCA to develop the subjective and objective qualities of the observed image, which result from filtering processes. Experiment results depict that the suggested algorithm performs efficiently over a wide range of visual contents and noise conditions, as well as in additive noise. Associated with different conventional noise estimators, the proposed algorithm yields the best performance, higher-quality images, and faster running speed.

All-zero-AC block detection using energy preservation theorem for H.263 video coding

This paper present a technique used in H.263 video coding to detect all-zero-AC occurrence without performing the transformation and quantization process. It is developed based on the energy preservation theorem. When the total AC energy in the pre-transform domain is known, the occurrence of all-zero-AC is predictable by comparing the total AC energy with the quantization threshold Q/sup 2/. The evaluations indicate that its prediction efficiency and speed gain depends on motion activities, level of Q and also the relative overhead imposed by the shortcut itself. In conclusion, the proposed new algorithm is expected to be practical in all common environments (low bit rate) targeted by H.263 standards as no negative speed gain is observed during the evaluation for all range of QP (Q=2.5QP).

Image edge detection operators based on orthogonal polynomials

ABSTRACT Orthogonal polynomials (OPs) are beneficial for image processing. OPs are used to reflect an image or a scene to a moment domain, and moments are subsequently used to extract object contours utilised in various applications. In this study, OP-based edge detection operators are introduced to replace traditional convolution-based and block processing methods with direct matrix multiplication. A mathematical model with empirical study results is established to investigate the performance of the proposed detectors compared with that of traditional algorithms, such as Sobel and Canny operators. The proposed operators are then evaluated by using entire images from a well-known data set. Experimental results reveal that the proposed operator achieves a more favourable interpretation, especially for images distorted by motion effects, than traditional methods do.

An extensive review on accessing quality information

The WWW has become one of the fastest growing electronic information sources. In the Web, people are engaging in interaction with more and more diverse information than ever before, so that the problem of information quality is more significant in the Web than any other information system, especially considering the rate of growth in the number of documents. Yet, despite a decade of active research, Information quality lacks comprehensive methodology for its assessment and improvement and a few researches have presented practical way for measuring IQ criteria. This paper attempts to address some of the issues involved in information quality assessment by classifying and discussing existing researches of information quality frameworks. While many of pervious works have concentrated on a dimension of information quality assessment, in our classification has been considered three dimensions: criteria, models and validation methods of the models and criteria. This classification prepares a bed for developing practical and at the same time valid information quality model.