Proceedings of International Conference on Image Processing
Rotation invariant texture descriptors are indispensable when the orientations of the training an... more Rotation invariant texture descriptors are indispensable when the orientations of the training and test images are different. Some of the well known rotation invariant texture analysis algorithms are based on the wavelet transform [4], the Laplacian pyramid filters [6], the ...
ABSTRACT This paper investigates the ability of a set of rotation invariant features to classify ... more ABSTRACT This paper investigates the ability of a set of rotation invariant features to classify images of wear particles found in used lubricating oil of machinery. The rotation invariant attribute of the features is derived from the property of the magnitudes of Fourier transform coefficients that do not change with spatial shift of the input elements. By analyzing individual circular neighborhoods centered at every pixel in an image, local and global texture characteristics of an image can be described. A number of input sequences are formed by the intensities of pixels on concentric rings of various radii measured from the center of each neighborhood. Fourier transforming the sequences would generate coefficients whose magnitudes are invariant to rotation. Rotation invariant features extracted from these coefficients were utilized to classify wear particle images that were obtained from a number of different particles captured at different orientations. In an experiment involving images of 6 classes, the circular neighborhood features obtained a 91% recognition rate which compares favorably to a 76% rate achieved by features of a 6 by 6 co-occurrence matrix.
This paper introduces a texture descriptor that is invariant to rotation. The new texture descrip... more This paper introduces a texture descriptor that is invariant to rotation. The new texture descriptor utilizes the property of the magnitudes of Fourier transform coefficients that do not change with spatial shift of input elements. Since rotating an image by an arbitrary angle does not change pixel intensities in an image but shift them in circular motion, the notion of producing textural features invariant to rotation using 1D Fourier transform coefficients can be realized if the relationship between circular motion and spatial shift can be established. By analyzing pixels in a circular neighborhood in an image, a number of FOurier transform coefficients can be generated to describe local properties of the neighborhood. From the magnitudes of these coefficients, several rotation invariant features are obtained to represent each texture class. Based on these features, an unknown image is assigned to one of the known classes using a nearest neighbor classifier. All of the feature samples for the classifier are extracted from unrotated texture images only. The new texture descriptor outperformed the circular simultaneous autoregressive model in classifying rotated texture images taken from 30 texture classes.
2016 IEEE International Conference on Power and Energy (PECon)
This paper prioritizes the different passive parameters on the basis of performance capability in... more This paper prioritizes the different passive parameters on the basis of performance capability in selecting the parameters for intelligent islanding detection techniques. The responses of 16 different passive parameters are analyzed under all possible islanding and non-islanding conditions. As a result of this, it has been found that rate of change of frequency over reactive power shows the highest capability to distinguish islanding from other events of similar signatures followed by rate of change of reactive power and rate of change of power.
This Provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted P... more This Provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted PDF and full text (HTML) versions will be made available soon. Automatic Cryptosporidium and Giardia viability detection in treated water
3rd Kuala Lumpur International Conference on Biomedical Engineering 2006
ABSTRACT This paper discusses on the classification of electrocardiogram (ECG) signal using multi... more ABSTRACT This paper discusses on the classification of electrocardiogram (ECG) signal using multiresolution wavelet transform and neural network. Multiresolution wavelet transform is used as a method of feature extraction of ECG signal since it has the ability to analyze the signal both in time and frequency domain. Neural network is used because of its ability to learn and perform classification on ECG signal. In this paper, four type of ECG signal has been chosen for classification. Based on the data obtained from MIT-BIH Arrhythmia Database the classification rate is found to be 95.08%.
2016 IEEE International Conference on Power and Energy (PECon), 2016
This paper prioritizes the different passive parameters on the basis of performance capability in... more This paper prioritizes the different passive parameters on the basis of performance capability in selecting the parameters for intelligent islanding detection techniques. The responses of 16 different passive parameters are analyzed under all possible islanding and non-islanding conditions. As a result of this, it has been found that rate of change of frequency over reactive power shows the highest capability to distinguish islanding from other events of similar signatures followed by rate of change of reactive power and rate of change of power.
A feature-based retinal image registration (RIR) technique aligns multiple fundus images and comp... more A feature-based retinal image registration (RIR) technique aligns multiple fundus images and composed of pre-processing, feature point extraction, feature descriptor, matching and geometrical transformation. Challenges in RIR include difference in scaling, intensity and rotation between images. The scale and intensity differences can be minimised with consistent imaging setup and image enhancement during the pre-processing, respectively. The rotation can be addressed with feature descriptor method that robust to varying rotation. Therefore, a feature descriptor method is proposed based on statistical properties (FiSP) to describe the circular region surrounding the feature point. From the experiments on public Fundus Image Registration dataset, FiSP established 99.227% average correct matches for rotations between 0° and 180°. Then, FiSP is paired with Harris corner, scale-invariant feature transform (SIFT), speeded-up robust feature (SURF), Ghassabi's and D-Saddle feature point extraction methods to assess its registration performance and compare with the existing feature-based RIR techniques, namely generalised dual-bootstrap iterative closet point (GDB-ICP), Harris-partial intensity invariant feature descriptor (PIIFD), Ghassabi's–SIFT, H-M 16, H-M 17 and D-Saddle–histogram of oriented gradients (HOG). The combination of SIFT–FiSP registered 64.179% of the image pairs and significantly outperformed other techniques with mean difference between 25.373 and 60.448% ( p = <0.001*).
ABSTRACT We demonstrate a passively multi-wavelength Q-switched Ytterbium-doped fiber laser (YDFL... more ABSTRACT We demonstrate a passively multi-wavelength Q-switched Ytterbium-doped fiber laser (YDFL) based on a multi-wall carbon nanotubes embedded in polyethylene oxide film as saturable absorber. The YDFL generates a stable multi-wavelength with spacing of 1.9 nm as the 980 nm pump power is fixed within 62. 4 mW and 78.0 mW. The repetition rate of the laser is tunable from 10.41 to 29.04 kHz by increasing the pump power from the threshold power of 62.4 mW to 78 mW. At 78 mW pump power, the maximum pulse energy of 38 nJ and the shortest pulse width of 8.87 µs are obtained.
Nanosecond pulse generation in an erbium-doped fiber laser (EDFL) passively mode-locked by a silv... more Nanosecond pulse generation in an erbium-doped fiber laser (EDFL) passively mode-locked by a silver nanoparticle (SNP)-based saturable absorber (SA) is experimentally demonstrated. The SA is fabricated by depositing a nanosized SNP layer onto the surface of polyvinyl alcohol film through the thermal evaporation process. By inserting the SA into an EDFL cavity, stable mode-locked operation is achieved at 1561.5 nm with the maximum pulse energy up to 52.3 nJ. The laser operates at a pulse repetition frequency of 1.0 MHz with a pulse width of 202 ns. These results suggest that SNPs could be developed as an effective SA for mode-locking pulse generation.
Q-switched and mode-locked fibre lasers have been successfully demonstrated in ytterbium-doped fi... more Q-switched and mode-locked fibre lasers have been successfully demonstrated in ytterbium-doped fibre laser cavity by taking an advantage of the optical absorption of antimony telluride (Sb2Te3) material. The Sb2Te3 was embedded into polyvinyl alcohol to function as a saturable absorber (SA). A Q-switching pulses train was obtained by incorporating the SA into the laser ring cavity configured with 3 dB coupler. The Q-switching pulse repetition rate increases from 24.4 to 55 kHz as the pump power is increased from the threshold of 75.4–96.2 mW. The maximum pulse energy of 252.6 nJ is obtained at 82.3 mW pump power. On the other hand, by changing the output coupler to 10 dB coupler, a stable self-started mode-locking operation was then generated at pump power range from 47.8 to 89.4 mW with a fixed repetition rate of 24.2 MHz. At 89.4 mW pump power, the maximum output power and pulse energy are obtained to be around 18.6 mW and 0.8 nJ, respectively. The authors results display that the Sb2Te3 material could also be developed as an effective SA for both Q-switched and mode-locked fibre lasers.
Brain electrical activity recordings by electroencephalography (EEG) are often contaminated with ... more Brain electrical activity recordings by electroencephalography (EEG) are often contaminated with signal artifacts. Procedures for automated removal of EEG artifacts are frequently sought for clinical diagnostics and brain computer interface (BCI) applications. In recent years, a combination of independent component analysis (ICA) and discrete wavelet transform (DWT) has been introduced as standard technique for EEG artifact removal. However, in performing the wavelet-ICA procedure, visual inspection or arbitrary thresholding may be required for identifying artifactual components in the EEG signal. We now propose a novel approach for identifying artifactual components separated by wavelet-ICA using a pre-trained support vector machine (SVM). Our method presents a robust and extendable system that enables fully automated identification and removal of artifacts from EEG signals, without applying any arbitrary thresholding. Using test data contaminated by eye blink artifacts, we show that our method performed better in identifying artifactual components than did existing thresholding methods. Furthermore, wavelet-ICA in conjunction with SVM successfully removed target artifacts, while largely retaining the EEG source signals of interest. We propose a set of features including kurtosis, variance, Shannon&#39;s entropy and range of amplitude as training and test data of SVM to identify eye blink artifacts in EEG signals. This combinatorial method is also extendable to accommodate multiple types of artifacts present in multi-channel EEG. We envision future research to explore other descriptive features corresponding to other types of artifactual components.
Proceedings of International Conference on Image Processing
Rotation invariant texture descriptors are indispensable when the orientations of the training an... more Rotation invariant texture descriptors are indispensable when the orientations of the training and test images are different. Some of the well known rotation invariant texture analysis algorithms are based on the wavelet transform [4], the Laplacian pyramid filters [6], the ...
ABSTRACT This paper investigates the ability of a set of rotation invariant features to classify ... more ABSTRACT This paper investigates the ability of a set of rotation invariant features to classify images of wear particles found in used lubricating oil of machinery. The rotation invariant attribute of the features is derived from the property of the magnitudes of Fourier transform coefficients that do not change with spatial shift of the input elements. By analyzing individual circular neighborhoods centered at every pixel in an image, local and global texture characteristics of an image can be described. A number of input sequences are formed by the intensities of pixels on concentric rings of various radii measured from the center of each neighborhood. Fourier transforming the sequences would generate coefficients whose magnitudes are invariant to rotation. Rotation invariant features extracted from these coefficients were utilized to classify wear particle images that were obtained from a number of different particles captured at different orientations. In an experiment involving images of 6 classes, the circular neighborhood features obtained a 91% recognition rate which compares favorably to a 76% rate achieved by features of a 6 by 6 co-occurrence matrix.
This paper introduces a texture descriptor that is invariant to rotation. The new texture descrip... more This paper introduces a texture descriptor that is invariant to rotation. The new texture descriptor utilizes the property of the magnitudes of Fourier transform coefficients that do not change with spatial shift of input elements. Since rotating an image by an arbitrary angle does not change pixel intensities in an image but shift them in circular motion, the notion of producing textural features invariant to rotation using 1D Fourier transform coefficients can be realized if the relationship between circular motion and spatial shift can be established. By analyzing pixels in a circular neighborhood in an image, a number of FOurier transform coefficients can be generated to describe local properties of the neighborhood. From the magnitudes of these coefficients, several rotation invariant features are obtained to represent each texture class. Based on these features, an unknown image is assigned to one of the known classes using a nearest neighbor classifier. All of the feature samples for the classifier are extracted from unrotated texture images only. The new texture descriptor outperformed the circular simultaneous autoregressive model in classifying rotated texture images taken from 30 texture classes.
2016 IEEE International Conference on Power and Energy (PECon)
This paper prioritizes the different passive parameters on the basis of performance capability in... more This paper prioritizes the different passive parameters on the basis of performance capability in selecting the parameters for intelligent islanding detection techniques. The responses of 16 different passive parameters are analyzed under all possible islanding and non-islanding conditions. As a result of this, it has been found that rate of change of frequency over reactive power shows the highest capability to distinguish islanding from other events of similar signatures followed by rate of change of reactive power and rate of change of power.
This Provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted P... more This Provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted PDF and full text (HTML) versions will be made available soon. Automatic Cryptosporidium and Giardia viability detection in treated water
3rd Kuala Lumpur International Conference on Biomedical Engineering 2006
ABSTRACT This paper discusses on the classification of electrocardiogram (ECG) signal using multi... more ABSTRACT This paper discusses on the classification of electrocardiogram (ECG) signal using multiresolution wavelet transform and neural network. Multiresolution wavelet transform is used as a method of feature extraction of ECG signal since it has the ability to analyze the signal both in time and frequency domain. Neural network is used because of its ability to learn and perform classification on ECG signal. In this paper, four type of ECG signal has been chosen for classification. Based on the data obtained from MIT-BIH Arrhythmia Database the classification rate is found to be 95.08%.
2016 IEEE International Conference on Power and Energy (PECon), 2016
This paper prioritizes the different passive parameters on the basis of performance capability in... more This paper prioritizes the different passive parameters on the basis of performance capability in selecting the parameters for intelligent islanding detection techniques. The responses of 16 different passive parameters are analyzed under all possible islanding and non-islanding conditions. As a result of this, it has been found that rate of change of frequency over reactive power shows the highest capability to distinguish islanding from other events of similar signatures followed by rate of change of reactive power and rate of change of power.
A feature-based retinal image registration (RIR) technique aligns multiple fundus images and comp... more A feature-based retinal image registration (RIR) technique aligns multiple fundus images and composed of pre-processing, feature point extraction, feature descriptor, matching and geometrical transformation. Challenges in RIR include difference in scaling, intensity and rotation between images. The scale and intensity differences can be minimised with consistent imaging setup and image enhancement during the pre-processing, respectively. The rotation can be addressed with feature descriptor method that robust to varying rotation. Therefore, a feature descriptor method is proposed based on statistical properties (FiSP) to describe the circular region surrounding the feature point. From the experiments on public Fundus Image Registration dataset, FiSP established 99.227% average correct matches for rotations between 0° and 180°. Then, FiSP is paired with Harris corner, scale-invariant feature transform (SIFT), speeded-up robust feature (SURF), Ghassabi's and D-Saddle feature point extraction methods to assess its registration performance and compare with the existing feature-based RIR techniques, namely generalised dual-bootstrap iterative closet point (GDB-ICP), Harris-partial intensity invariant feature descriptor (PIIFD), Ghassabi's–SIFT, H-M 16, H-M 17 and D-Saddle–histogram of oriented gradients (HOG). The combination of SIFT–FiSP registered 64.179% of the image pairs and significantly outperformed other techniques with mean difference between 25.373 and 60.448% ( p = <0.001*).
ABSTRACT We demonstrate a passively multi-wavelength Q-switched Ytterbium-doped fiber laser (YDFL... more ABSTRACT We demonstrate a passively multi-wavelength Q-switched Ytterbium-doped fiber laser (YDFL) based on a multi-wall carbon nanotubes embedded in polyethylene oxide film as saturable absorber. The YDFL generates a stable multi-wavelength with spacing of 1.9 nm as the 980 nm pump power is fixed within 62. 4 mW and 78.0 mW. The repetition rate of the laser is tunable from 10.41 to 29.04 kHz by increasing the pump power from the threshold power of 62.4 mW to 78 mW. At 78 mW pump power, the maximum pulse energy of 38 nJ and the shortest pulse width of 8.87 µs are obtained.
Nanosecond pulse generation in an erbium-doped fiber laser (EDFL) passively mode-locked by a silv... more Nanosecond pulse generation in an erbium-doped fiber laser (EDFL) passively mode-locked by a silver nanoparticle (SNP)-based saturable absorber (SA) is experimentally demonstrated. The SA is fabricated by depositing a nanosized SNP layer onto the surface of polyvinyl alcohol film through the thermal evaporation process. By inserting the SA into an EDFL cavity, stable mode-locked operation is achieved at 1561.5 nm with the maximum pulse energy up to 52.3 nJ. The laser operates at a pulse repetition frequency of 1.0 MHz with a pulse width of 202 ns. These results suggest that SNPs could be developed as an effective SA for mode-locking pulse generation.
Q-switched and mode-locked fibre lasers have been successfully demonstrated in ytterbium-doped fi... more Q-switched and mode-locked fibre lasers have been successfully demonstrated in ytterbium-doped fibre laser cavity by taking an advantage of the optical absorption of antimony telluride (Sb2Te3) material. The Sb2Te3 was embedded into polyvinyl alcohol to function as a saturable absorber (SA). A Q-switching pulses train was obtained by incorporating the SA into the laser ring cavity configured with 3 dB coupler. The Q-switching pulse repetition rate increases from 24.4 to 55 kHz as the pump power is increased from the threshold of 75.4–96.2 mW. The maximum pulse energy of 252.6 nJ is obtained at 82.3 mW pump power. On the other hand, by changing the output coupler to 10 dB coupler, a stable self-started mode-locking operation was then generated at pump power range from 47.8 to 89.4 mW with a fixed repetition rate of 24.2 MHz. At 89.4 mW pump power, the maximum output power and pulse energy are obtained to be around 18.6 mW and 0.8 nJ, respectively. The authors results display that the Sb2Te3 material could also be developed as an effective SA for both Q-switched and mode-locked fibre lasers.
Brain electrical activity recordings by electroencephalography (EEG) are often contaminated with ... more Brain electrical activity recordings by electroencephalography (EEG) are often contaminated with signal artifacts. Procedures for automated removal of EEG artifacts are frequently sought for clinical diagnostics and brain computer interface (BCI) applications. In recent years, a combination of independent component analysis (ICA) and discrete wavelet transform (DWT) has been introduced as standard technique for EEG artifact removal. However, in performing the wavelet-ICA procedure, visual inspection or arbitrary thresholding may be required for identifying artifactual components in the EEG signal. We now propose a novel approach for identifying artifactual components separated by wavelet-ICA using a pre-trained support vector machine (SVM). Our method presents a robust and extendable system that enables fully automated identification and removal of artifacts from EEG signals, without applying any arbitrary thresholding. Using test data contaminated by eye blink artifacts, we show that our method performed better in identifying artifactual components than did existing thresholding methods. Furthermore, wavelet-ICA in conjunction with SVM successfully removed target artifacts, while largely retaining the EEG source signals of interest. We propose a set of features including kurtosis, variance, Shannon&#39;s entropy and range of amplitude as training and test data of SVM to identify eye blink artifacts in EEG signals. This combinatorial method is also extendable to accommodate multiple types of artifacts present in multi-channel EEG. We envision future research to explore other descriptive features corresponding to other types of artifactual components.
Uploads
Papers by Hamzah Arof