Atif Bin Mansoor

A feature level multimodal approach for palmprint identification using directional subband energies

Palmprint based personal identification has gained preference over other biometric modalities due to its ease of acquisition, high user acceptance and reliability. This paper presents a palmprint based identification approach which uses the textural information available on the palmprint by employing a feature level fusion of contourlet transform (CT) and non-subsampled contourlet transform (NSCT). The proposed algorithm captures both local and global details in a palmprint as a compact fixed length palm code. After establishing the region of interest (ROI), the two-dimensional (2-D) spectrum is divided into fine slices using iterated directional filterbanks. Next, directional energy component for each block from the decomposed subband outputs is computed separately for the two transforms. The features from both domains are then fused at feature levels. Palmprint matching is then performed using normalized Euclidean distance classifier. The algorithm is tested on complete database of 7752 palm images of Polytechnic University of Hong Kong, and 500 palm images of GPDS Hand database from University of Las Palmas de Gran Canaria. The experimental results were compiled for features based upon individual transforms and fused one. CT based approach demonstrated the decidability index of 2.7734 and equal error rate (EER) of 0.2333% while NSCT based approach depicted decidability index of 2.8125 and EER of 0.1604% on palm database of Polytechnic University of Hong Kong. Similarly, CT based approach demonstrated the decidability index of 2.6212 and equal error rate (EER) of 0.7082% while NSCT based approach depicted decidability index of 2.7278 and EER of 0.5082% on GPDS hand database. The multimodal approach based upon feature fusion achieved decidability index of 2.8914 and EER of 0.1563% on database of Polytechnic University of Hong Kong and decidability index of 2.7956 and EER of 0.3112% on GPDS hand database. The quantitative measures confirm progressive improved results in three approaches for both the databases.

Combined Contourlet and Non-subsampled Contourlet Transforms Based Approach for Personal Identification Using Palmprint

Palmprint based personal verification is an accepted biometric modality due to its reliability, ease of acquisition and user acceptance. This paper presents a novel palmprint based identification approach which draw on the textural information available on the palmprint by utilizing a combination of Contourlet and Non Subsampled Contourlet Transforms. Center of the palm is computed using the Distance Transform whereas the parameters of best fitting ellipse help determine the alignment of the palmprint. ROI of 256X256 pixels is cropped around the center, and subsequently it is divided into fine slices, using iterated directional filterbanks. Next, directional energy components for each block of the decomposed subband outputs are computed using Contourlet and Non Subsampled Contourlet Transforms. The proposed algorithm captures global details in a palmprint as compact fixed length palm codes for Contourlet and NSCT respectively which are further concatenated at feature level for identification using normalized Euclidean distance classifier. The proposed algorithm is tested on a total of 500 palm images of GPDS Hand database, acquired from University of Las Palmas de Gran Canaria. The experimental results were compiled for individual transforms as well as for their optimized combination at feature level. CT based approach demonstrated the Decidability Index of 2.6212 and Equal Error Rate (EER) of 0.7082% while NSCT based approach depicted Decidability Index of 2.7278 and EER of 0.5082%. The feature level fusion achieved Decidability Index of 2.7956 and EER of 0.3112%.

Fuzzy morphology for edge detection and segmentation

This paper proposes a new approach for structure based separation of image objects using fuzzy morphology. With set operators in fuzzy context, we apply an adaptive alpha-cut morphological processing for edge detection, image enhancement and segmentation. A Top-hat transform is first applied to the input image and the resulting image is thresholded to a binary form. The image is then thinned using hit-or-miss transform. Finally, m-connectivity is used to keep the desired number of connected pixels. The output image is overlayed on the original for enhanced boundaries. Experiments were performed using real images of aerial views, sign boards and biological objects. A comparison to other edge enhancement techniques like unsharp masking, sobel and laplacian filtering shows improved performance by the proposed technique.

Personal identification using feature and score level fusion of palm- and fingerprints

The ever increasing demand of security has resulted in wide use of Biometric systems. Despite overcoming the traditional verification problems, the unimodal systems suffer from various challenges like intra class variation, noise in the sensor data etc, affecting the system performance. These problems are effectively handled by multimodal systems. In this paper, we present multimodal approach for palm- and fingerprints by feature level and score level fusions (sum and product rules). The proposed multi-modal systems are tested on a developed database consisting of 440 palm- and fingerprints each of 55 individuals. In feature level fusion, directional energy-based feature vectors of palm- and fingerprint identifiers are combined to form joint feature vector that is subsequently used to identify the individual using a distance classifier. In score level fusion, the matching scores of individual classifiers are fused by sum and product rules. Receiver operating characteristics curves are formed for unimodal and multimodal systems. Equal Error Rate (EER) of 0.538% for feature level fusion shows best performance compared to score level fusion of 0.6141 and 0.5482% of sum and product rules, respectively. Multimodal systems, however, significantly outperform unimodal palm- and fingerprints identifiers with EER of 2.822 and 2.553%, respectively.

On Image Compression using Digital Curvelet Transform

This paper describes a novel approach to digital image compression using a new mathematical transform: the curvelet transform. The transform has shown promising results over wavelet transform for 2D signals. Wavelets, though well suited to point singularities have limitations with orientation selectivity, and therefore, do not represent two-dimensional singularities (e.g. smooth curves) effectively. This paper employs the curvelet transform for image compression, exhibiting good approximation properties for smooth 2D functions. Curvelet improves wavelet by incorporating a directional component. The curvelet transform finds a direct discrete-space construction and is therefore computationally efficient. In this paper, we divided 2D spectrum into fine slices using iterated tree structured filter bank. Different amount of quantized curvelet coefficients were then selected for lossy compression and entropy encoding. A comparison with wavelet based compression was made for standard images like Lena, Barbara, etc. Curvelet transform has resulted in high quality image compression for natural images. Our implementation offers exact reconstruction, prone to perturbations, ease of implementation and low computational complexity. The algorithm works fairly well for grayscale and colored images

Directional energy based palmprint identification using Non Subsampled Contourlet Transform

Palmprint based personal verification has gained preference over other biometric modalities due to its ease of acquisition, high user acceptance and reliability. This paper presents a novel palmprint based identification approach which uses the textural information available on the palmprint by employing the Non Subsampled Contourlet Transform (NSCT). After establishing the region of interest (ROI), the two dimensional (2-D) spectrum is divided into fine slices, using iterated directional filterbanks. Next, directional energy component for each block from the decomposed subband outputs is computed. The proposed algorithm captures both local and global details in a palmprint as a compact fixed length palm code. Palmprint matching is then performed using Normalized Euclidean Distance classifier. The algorithm is tested on a total of 7752 palm images, acquired from the standard database of Polytechnic University of Hong Kong. The experimental results demonstrated the feasibility of the proposed system by exhibiting Decidability Index of 2.8125 and Equal Error Rate of 0.1604%, better than the reported techniques in literature.

Performance comparison of two altitude-control algorithms for a fixed-wing UAV

Uninhabited aerial vehicles (UAV) have proven their tremendous capabilities in military and civil applications. In a UAV, the onboard autopilot autonomously controls the aircraft flight and navigation. The altitude acquire-and-hold is an important function of autopilot, implemented using a control design algorithm that flies the UAV to commanded altitude and maintains it. Most of the commercially available autopilots use Proportional-Integral-Derivative (PID) controllers for altitude and heading control. In this paper, we present a performance comparison of two altitude-controller design techniques, the PID controller and the Phase Lead compensator. We have used a nonlinear mathematical model of the UAV Aerosonde in our work. The nonlinear model is lineraized around a stable trim condition and decoupled for linear controller design. The designed controllers are tested with the nonlinear model in view of small perturbation control theory. The results for the compensated linear and nonlinear models are presented. Our investigation reveals that Phase Lead compensator has inherent strengths compared to PID controller for UAV altitude acquire- and-hold in terms of better transient response, thus improving the payload performance during an altitude-change maneuver. The findings may lead to an effective approach in UAV autopilot design.

A New Approach to Aircraft Surface Inspection Based on Directional Energies of Texture

Non Destructive Inspections (NDI) plays a vital role in aircraft industry as it determines the structural integrity of aircraft surface and material characterization. The existing NDI methods are time consuming, we propose a new NDI approach using Digital Image Processing that has the potential to substantially decrease the inspection time. The aircraft imagery is analyzed by two methods i.e Contourlet Transform (CT) and Discrete Cosine Transform (DCT). With the help of Contourlet Transform the two dimensional (2-D) spectrum is divided into fine slices, using iterated directional filter banks. Next, directional energy components for each block of the decomposed subband outputs are computed. These energy values are used to distinguish between the crack and scratch images using the Dot Product classifier. In next approach, the aircraft imagery is decomposed into high and low frequency components using DCT and the first order moment is determined to form feature vectors. A correlation based approach is then used for distinction between crack and scratch surfaces. A comparative examination between the two techniques on a database of crack and scratch images revealed that texture analysis using the combined transform based approach gave the best results by giving an accuracy of 96.6% for the identification of crack surfaces and 98.3% for scratch surfaces.

Image processing and classification in diabetic retinopathy: A review

Diabetic retinopathy is one of the disabling microvascular complications of diabetes mellitus that causes the loss of central vision or in cases complete vision loss if not recognized and cured at the earlier stage. This work reviews the latest techniques in digital image processing and pattern classification employed for the detection of diabetic retinopathy and compares them on the basis of different performance measures like sensitivity, specificity, accuracy and area under the curve in receiver operating characteristic. The classification of diabetic retinopathy follows various steps like pre-processing, feature extraction and classification of microaneurysm, hemorrhages, exudates and cotton woolen spot. In this paper, the reported literature in each domain is analyzed.

A novel contourlet based online fingerprint identification

Biometric based personal identification is regarded as an effective method for automatically recognizing an individuals identity. As a method for preserving the security of sensitive information biometrics has been applied in various fields over last few decades. In our work, we present a novel core based global matching approach for fingerprint matching using the Contourlet Transform. The core and delta points along with the ridge and valley orientations have strong directionality or directional information. This directionality has been exploited as the features and considered for matching. The obtained ROI is analyzed for its textures using Contourlet transform which divides the 2-D spectrum into fine slices by employing Directional Filter Banks (DFBs). Distinct features are then extracted at different resolutions by calculating directional energies for each sub-block from the decomposed subband outputs, and given to a Euclidian distance classifier. Finally adaptive majority vote algorithm is employed in order to further narrow down the matching criterion. The algorithm has been tested on a developed database of 126 individuals, enrolled with 8 templates each.