Electronic Letters on Computer Vision and Image Analysis (ELCVIA - Universitat Autònoma de Barcelona)
Not a member yet
343 research outputs found
Sort by
Robust human detection through fusion of color and infrared video
Get PDFThis PhD thesis develops and implements a robust multisensor human detection systembased on fusing the information provided after segmenting infrared and color videos
Colour constancy in natural images through colour naming and sensor sharpening
Get PDFColour is derived from three physical properties: incident light, object reflectance and sensor sensitivities. Incident light varies under natural conditions; hence, recovering scene illuminant is an important issue in computational colour. One way to deal with this problem under calibrated conditions is by following three steps, 1) building a narrow-band sensor basis to accomplish the diagonal model, 2) building a feasible set of illuminants, and 3) defining criteria to select the best illuminant. In this work we focus on colour constancy for natural images by introducing perceptual criteria in the first and third stages.To deal with the illuminant selection step, we hypothesize that basic colour categories can be used as anchor categories to recover the best illuminant. These colour names are related to how the human visual system has evolved to encode relevant natural colour statistics. Therefore the recovered image provides the best representation of the scene labelled with the basic colour terms. We demonstrate with several experiments how this selection criterion achieves current state-of-art results in computational colour constancy. In addition to this result, we psychophysically prove that usual angular error used in colour constancy does not correlate with human preferences, and we propose a new perceptual colour constancy evaluation.The implementation of this selection criterion strongly relies on the use of a diagonal model for illuminant change. Then, the second contribution focuses on building an appropriate narrow-band sensor basis to represent natural images. We propose to use the spectral sharpening technique to compute a unique narrow-band basis optimized to represent a large set of natural reflectances under natural illuminants and given in the basis of human cones. The proposed sensors allow predicting unique hues and the World colour Survey data independently of the illuminant by using a compact singularity function. Additionally, we studied different families of sharp sensors to minimize different perceptual measures. This study brought us to extend the spherical sampling procedure from 3D to 6D.Several research lines remain still open, such as, measuring the effects of using the computed sharp sensors on the category hypothesis; or inserting spatial contextual information to improve category hypothesis. Finally, to explore how individual sensors can be adjusted to the colours in a scene
Blind Restoration of Motion Blurred Barcode Images using Ridgelet Transform and Radial Basis Function Neural Network
Get PDFThe aim of any image restoration techniques is recovering the original image from a degraded observation. One of the most common degradation phenomena in images is motion blur. In case of blind image restoration accurate estimation of motion blur parameters is required for deblurring of such images. This paper proposed a novel technique for estimating the parameters of motion blur using ridgelet transform. Initially, the energy of ridgelet coefficients is used to estimate the blur angle and then blur length is estimated using a radial biases function neural network. This work is tested on different barcode images with varying parameters of blur. The simulation results show that the proposed method improves the restoration performance
Semantic Awareness for Automatic Image Interpretation
Get PDFFinding relations between image semantics and image characteristics is a problem of long standing in computer vision, image analysis or related fields. Classic research in these fields is intended for applications that go from the image domain to the semantic domain such as face recognition or scene understanding. This thesis explores methods and applications that go the opposite direction, i.e. use existing semantic information to infer knowledge and actions in the image domain. We build a large scale statistical framework to relate image characteristics to semantic expressions for millions of images and thousands of keywords. We apply the framework to semantic image enehancement and automatic color naming
Autonomous UAV for Suspicious Action Detection using Pictorial Human Pose Estimation and Classification
Get PDFVisual autonomous systems capable of monitoring crowded areas and alerting the authorities in occurrence of a suspicious action can play a vital role in controlling crime rate. Previous atte mpts have been made to monitor crime using posture recognition but nothing exclusive to investigating actions of people in large populated area has been cited. In order resolve this shortcoming, we propose an autonomous unmanned aerial vehicle (UAV) visual surveillance system that locates humans in image frames followed by pose estimation using weak constraints on position, appearance of body parts and image parsing. The estimated pose, represented as a pictorial structure, is flagged using the proposed Hough Orientation Calculator (HOC) on close resemblance with any pose in the suspicious action dataset. The robustness of the system is demonstrated on videos recorded using a UAV with no prior knowledge of background, lighting or location and scale of the human in the image. The system produces an accuracy of 71% and can also be applied on various other video sources such as CCTV camera
SWT voting-based color reduction method for detecting text in natural scene images
Get PDFIn our PhD thesis we give a very detailed and in-depth survey of natural scene text detection methods and propose two novel methods, namely SWT (Stroke Width Transform) voting-based color reduction method and SWT direction determination method. SWT voting-based color reduction method (to which we will refer also as SWT-V) is a novel text detection method that - opposed to many other text detection methods - combines both structural and color information in order to detect text. The proposed method upgrades the text detection oriented color reduction method (to which we will refer to as TOCR) with the additional SWT voting stage and substantially outperforms other state-of-the-art text detection methods. All the image colors rich with SWT pixels that most likely belong to text characters are blocked from being mean-shifted away in the color reduction process. One of the disadvantages of the SWT method, however, is the problem of ‘light text on the dark background’ described in the following sections. To cope with the problem and in order to provide true SWT values to the SWT voting stage we propose an adaptive SWT direction determination method. The method uses SWT profiles to partition an image into subblocks and analyzes their SWT histograms of both SWT search directions. Text detection literature does not explicitly address the SWT direction issue, therefore, the proposed method represents a unique scientific contribution to the research field. All text detection methods were evaluated on the CVL OCR DB text detection evaluation dataset
Multitier Biometric Template Security Using Cryptographic Salts and Personal Image Identification
Get PDFIndividual identification can be accurately done by measuring biological parameters termed as biometrics. These have been proved as an exceptional tool for identity verification. Security of biometric template is the most challenging aspect of biometric identification system. Storing the biometric template in the database increases the chance of compromising it which may lead to serious threat and misuse of the individual identity. This paper proposes a novel and computationally simpler approach to store a biometric sample in the form of template by using cryptographic salts. Use of Personal Image Identification (PII) makes the proposed algorithm more robust and adds another level of security. The saltcrypted templates are created and stored instead of storing the actual sample behaving as a fuzzy vault. The algorithm has been analytically proved computationally simple compared to the existing template security mechanisms. The fuzzy structure of saltcrypted template is entirely dependent on user interaction through PII. Actual template is not stored at any point of time which adds new dimension to the security and hence to individual identity
Detecting and indexing moving objects for Behavior Analysis by Video and Audio Interpretation
Get PDFIn this thesis a system for analyzing moving objects behavior for surveillance applications is proposed: videos are processed in order to extract and analyze moving objects trajectories for identifying abnormal trajectories, associated to abnormal behaviors. Whereas the information extracted from the videos are not sufficient or not sufficiently reliable, the proposed system is enriched by a module in charge of recognizing audio events of interest such as shoots, screams or broken glasses. Finally, all the extracted information are suitably stored in order to allow an efficient retrieval from the human operator. Five different standard datasets have been used for testing the different modules proposed in this thesis; the obtained results, both in terms of accuracy and computational efficiency, confirm the effectiveness and the real applicability of the proposed approach
FMIRS: A Fuzzy indexing and retrieval system of mosaic-image database
Get PDFThis work is dedicated to present a fuzzy-set based system useful for image indexing and retrieval pertaining to historical Roman-mosaics. This exceptional collection of mosaics dates back from the first to fourth centuries AD. Considering the state of these images (i.e. noise, color degradation, etc.) a fuzzy features definition is necessary. Thereby, we use a robust to rotation, scale and translation fuzzy extended curvature scale space (CSS) as shape descriptor. Furthermore, we propose a fuzzy color-quantization approach, applied on mosaics, using HSV color space. The system allows for two user-friendly querying modes: a drawing based mode and the mode that fusion both shape and color features using a unified fuzzy similarity measure. Based on queries of variable complexity, the advanced fuzzy system has managed to achieve interesting recall, precision and F-measure rates