Electronic Letters on Computer Vision and Image Analysis (ELCVIA - Universitat Autònoma de Barcelona)
Not a member yet
343 research outputs found
Sort by
An Automatic Ship Detection Method Based on Local Gray-Level Gathering Characteristics in SAR Imagery
Get PDFThis paper proposes an automatic ship detection method based on gray-level gathering characteristics of synthetic aperture radar (SAR) imagery. The method does not require any prior knowledge about ships and background observation. It uses a novel local gray-level gathering degree (LGGD) to characterize the spatial intensity distribution of SAR image, and then an adaptive-like LGGD thresholding and filtering scheme to detect ship targets. Experiments on real SAR images with varying sea clutter backgrounds and multiple target situations have been conducted. The performance analysis confirms that the proposed method works well in various circumstances with high detection rate, fast detection speed and perfect shape preservation
Peak Trekking of Hierarchy Mountain for the Detection of Cerebral Aneurysm using Modified Hough Circle Transform
Get PDFThe Circle of Willis is the junction of two carotid arteries and two vertebral arteries that supply the brain with nutrition. Junctions where these arteries come together may develop weak spots that can balloon out and fill with blood, creating aneurysms. These sac-like areas may leak or rupture, spilling blood into surrounding tissues which may cause artery spasm leading to potential stroke or even death. Clipping and coiling are two treatment options preferred by neurosurgeon which require proper detection of aneurysm. Medical practitioners are therefore emphasizing on the prior detection of cerebral aneurysm (CA) before rupture occurs leading to subarachnoid haemorrhage (SAH). This paper presents an efficient detection technique rooted in Modified Hough Circle Transform (MHCT) on the image extracted from Digital Subtraction Angiogram (DSA). Experimental results have firmly substantiated that the proposed method is highly efficient in properly detecting the location, size and type of aneurysm
Image based Monument Recognition using Graph based Visual Saliency
Get PDFThis article presents an image-based application aiming at simple image classification of well-known monumentses in the area of Heraklion, Crete, Greece. This classification takes place by utilizing Graph Based Visual Saliency (GBVS) and employing Scale Invariant Feature Transform (SIFT) or Speeded Up Robust Features (SURF). For this purpose, images taken at various places of interest are being compared to an existing database containing images of these places at different angles and zoom. The time required for the matching progress in such application is an important element. To this goal, the images have been previously processed according to the Graph Based Visual Saliency model in order to keep either SIFT or SURF features corresponding to the actual monuments while the background “noise” is minimized. The application is then able to classify these images, helping the user to better understand what he/she sees and in which area he had his/her image taken. Experiments are performed to verify that the proposed approach improves the time needed for the classification without affecting the correctness of the results
An Interactive Deformable Model Segmentation Algorithm Driven by Morphological Dilations and Erosions Constrained by an Exclusion Band
Get PDFThis study introduces an interactive image segmentation algorithm for extraction of ill-defined edges (faint, blurred or partially broken) often observed at small-scale imaging. It is based on a simplified deformable elastic model evolution paradigm. Segmentation is achieved as a two-step region-growing, shrinking and merging simulation constrained by an exclusion band built around the edges of the regions of interest, defined from a variation image. The simulation starts from a set of unlabeled markers and the respective elastic models. During the first step, model evolution occurs entirely outside the exclusion band, driven by alternate action-reaction movements. Forward and backward movements are performed by constrained binary morphological dilations and erosions. Constraints allow controlling how far models can move through narrow gaps. At the end of the first step, models remaining from merging operations receive unique and exclusive labels. On the second and final step, models expansion occurs entirely inside the exclusion band, now driven only by binary unconstrained morphological dilations. A point where two labeled models get into contact defines an edge point. The simulation goes on until the concurrent expansion of all models comes to a complete stop. At this point, the edges of the regions-of-interest have been extracted. Interactivity introduces the possibility to correct small imperfections in the edge positioning by changing a parameter controlling action-reaction or by changing marker’s size, position and shape. Slightly inspired by traditional approaches as PDE Level-Set based curve evolution and Immersion Simulation, the algorithm presents a solution to the problem of “synchronizing the concurrent evolution of a large number of models” and an “automatic stopping criterion” for the front propagation. Integer arithmetic implementation assures linear execution time. The results obtained for real applications show that even ill-defined edges can be located with the desired accuracy, thanks to algorithm features and to the interactivity exerted by the user during the segmentation procedure
Novel CBIR system based on Ripplet Transform using interactive Neuro-Fuzzy technique
Get PDFContent Based Image Retrieval (CBIR) system is an emerging research area in effective digital data management paradigm. In this article, a novel CBIR system based on a new Multiscale Geometric Analysis (MGA)-tool, called Ripplet Transform Type-I (RT) is presented. To improve the retrieval result and to decrease the computational cost, the proposed scheme consists of a Neural Network (NN) based classifier for image pre-classification, similarity matching using Manhattan distance measure and relevance feedback mechanism (RFM) using fuzzy entropy based feature evaluation technique. Extensive experiments were carried out to evaluate the effectiveness of the proposed scheme. Experimental results and comparisons show that the proposed CBIR system performs efficiently
Implementing Cepstral Filtering Technique using Gabor Filters
Get PDFCepstral filtering technique is applied on an interlaced image, the pattern similar to that which is found in layer IV of Primate Visual Cortex. It involves Power spectrum in computation, which is square of absolute of Fast Fourier Transform (FFT), is a complicated and hardware unfriendly. We propose an algorithm in which Gabor filters, instead of Power Spectrum, are applied to an interlaced image in the Cepstral algorithm. This scheme makes it hardware friendly as it gives the flexibility of working with modules which can be imitated in hardware. Building a FFT module is a tough task in analog circuit but determining Gabor energy, an alternative to it, can be achieved by elementary circuits. The Phase, Energy Models and other methods, use multi-lambda Gabor filters to compute disparity. The proposed method uses sum of absolute difference to choose a single Gabor filter of appropriate lambda that fits to find the disparity. The algorithm inherits the quality of both Gabor filter and Ocular Dominance Pattern and hence a biologically inspired and suitable for hardware realization. The proposed algorithm has been implemented on the test data image. A hardware scheme has also been proposed that can be used to estimate disparity and the idea can be extended in building complex modules that can perform real time - real image operations with a handful of resources as compared to employing complex digital FPGAs and CPLDs
Handwritten Digit Recognition by Fourier-Packet Descriptors
Get PDFAny statistical pattern recognition system includes a feature extraction component. For character patterns, several feature families have been tested, such as the Fourier-Wavelet Descriptors. We are proposing here a generalization of this family: the Fourier-Packet Descriptors. We have tested a set of 72 of these features on handwritten digits: the error rate was 2.44% with classifier 1NN for 19 features selected from the set and 1.72% with classifier SVM for all the set
Facial Expression Recognition Using New Feature Extraction Algorithm
Get PDFThis paper proposes a method for facial expression recognition. Facial feature vectors are generated from keypoint descriptors using Speeded-Up Robust Features. Each facial feature vector is then normalized and next the probability density function descriptor is generated. The distance between two probability density function descriptors is calculated using Kullback Leibler divergence. Mathematical equation is employed to select certain practicable probability density function descriptors for each grid, which are used as the initial classification. Subsequently, the corresponding weight of the class for each grid is determined using a weighted majority voting classifier. The class with the largest weight is output as the recognition result. The proposed method shows excellent performance when applied to the Japanese Female Facial Expression database
Vision based Object Recognition and Localization by a Network Connected Distributed Robotic Systems
No full textObject recognition and Localization is an important problem in computer vision and robotics. Advances in computer vision has resulted into many object recognition techniques but most of them are computationally very heavy and requires robot unit to have high processing systems. When it comes to small size embedded robotic systems, these techniques can not be applied because of the constraints of execution time. Here, a popular SURF based recognition technique is adopted and some of the important changes are addressed which make it possible to use it on small size robots with limited resources. A team of small robots are used which are given prior training to search for 2D and 3D objects of interest in the environment. For the localization of the robots and objects a new template, designed for passive markers based tracking, is introduced. These markers are placed on the top of each robot and they are tracked by the two ceiling mounted cameras. The information from ceiling mounted cameras and from the team of robots is used collectively to localize the object of interest in the environment
Fuzzy Binary Patterns for uncertainty-aware texture representation
Get PDFThe Local Binary Pattern (LBP) representation of textures has been proved useful for a wide range of pattern recognition applications, including texture segmentation, face detection, and biomedical image analysis. The interest of the research community in the LBP texture representation gave rise to plenty of LBP and other binary pattern (BP)-based variations. However, noise sensitivity is still a major concern to their applicability on the analysis of real world images. To cope with this problem we propose a generic, uncertainty-aware methodology for the derivation of Fuzzy BP (FBP) texture models. The proposed methodology assumes that a local neighbourhood can be partially characterized by more than one binary patterns due to noise-originated uncertainty in the pixel values. The texture discrimination capability of four representative FBP-based approaches has been evaluated on the basis of comprehensive classification experiments on three reference datasets of natural textures under various types and levels of additive noise. The results reveal that the FBP-based approaches lead to consistent improvement in texture classification as compared with the original BP-based approaches for various degrees of uncertainty. This improved performance is also validated by illustrative unsupervised segmentation experiments on natural scenes