Search results for "computer vision"
showing 10 items of 2353 documents
Physical simulation for real-time image/video retargeting
2012
Retargeting methods present content on arbitrary aspect ratio me- dia displays limiting distortions in relevant objects. This is done by means of applying non-homogeneous resizing operators across the whole media, constraining it to fit into the required size. Several succesful systems have been proposed to achieve image retargeting, while video retargeting is still challenging due to time consistency and computational complexity requirements. Two important con- tributes were proposed: non-homogeneous retargeting [Wolf et al. 2007] and improved seam-carving [Rubinstein et al. 2008]. The first one claims to achieve real-time performance but considers spa- tial coordinates separately. The lat…
Using the fibre structure of paper to determine authenticity of the documents: analysis of transmitted light images of stamps and banknotes.
2014
A novel method is presented for distinguishing postal stamp forgeries and counterfeit banknotes from genuine samples. The method is based on analyzing differences in paper fibre networks. The main tool is a curvelet-based algorithm for measuring overall fibre orientation distribution and quantifying anisotropy. Using a couple of more appropriate parameters makes it possible to distinguish forgeries from genuine originals as concentrated point clouds in two- or three-dimensional parameter space.
Including invariances in SVM remote sensing image classification
2012
This paper introduces a simple method to include invariances in support vector machine (SVM) for remote sensing image classification. We rely on the concept of virtual support vectors, by which the SVM is trained with both the selected support vectors and synthetic examples encoding the invariance of interest. The algorithm is very simple and effective, as demonstrated in two particularly interesting examples: invariance to the presence of shadows and to rotations in patchbased image segmentation. The improved accuracy (around +6% both in OA and Cohen's κ statistic), along with the simplicity of the approach encourage its use and extension to encode other invariances and other remote sensin…
Extraction of objects from structured backgrounds in the cat superior colliculus. Part II
1980
Specific changes occur in the cells of the uppers layers of the cat's superior colliculus when a two dimensional noise (background) is superimposed onto a deterministic signal (spot of light). Some of the measurements can be interpreted as meaning that some cells only react to certain relative movements of object (spot) and background (noise). The movement of the visual background is interpreted as environmental movement occurring due to the animal's own movement. The results of the measurements provide all the necessary presuppositions for a distinction between the animal's own velocity and that of the object (Part I). The experimental results can be interpreted with a model. The essential…
Rotational panoramic radiographs-unusual triple images
2015
Currently clinicians advice rotational panoramic radiography (RPR) for preliminary investigation. Despite few inherent limitations, rotational panoramic radiography still remains the diagnostic tool of choice. Abnormal structures such as a supernumerary tooth or a device falling within the certain central regions in conventional RPR images may mislead the clinicians towards an inaccurate diagnosis by producing multiple ghost images. Such cases must be treated with circumspect, and apart from RPR, additional imaging modalities need be employed to provide a judicious interpretation of the clinical situation. Thus this manuscript, we present a case where a paramedian supernumerary tooth which …
Classification and Automated Interpretation of Spinal Posture Data Using a Pathology-Independent Classifier and Explainable Artificial Intelligence (…
2021
Clinical classification models are mostly pathology-dependent and, thus, are only able to detect pathologies they have been trained for. Research is needed regarding pathology-independent classifiers and their interpretation. Hence, our aim is to develop a pathology-independent classifier that provides prediction probabilities and explanations of the classification decisions. Spinal posture data of healthy subjects and various pathologies (back pain, spinal fusion, osteoarthritis), as well as synthetic data, were used for modeling. A one-class support vector machine was used as a pathology-independent classifier. The outputs were transformed into a probability distribution according to Plat…
Classification of SD-OCT volumes with multi pyramids, LBP and HOG descriptors: application to DME detections.
2016
This paper deals with the automated detection of Diabetic Macular Edema (DME) on Optical Coherence Tomography (OCT) volumes. Our method considers a generic classification pipeline with preprocessing for noise removal and flattening of each B-Scan. Features such as Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP) are extracted and combined to create a set of different feature vectors which are fed to a linear-Support Vector Machines (SVM) Classifier. Experimental results show a promising sensitivity/specificity of 0.75/0.87 on a challenging dataset.
An architecture for automatic gesture analysis
2000
The field of human-computer interaction has been widely investigated in the last years, resulting in a variety of systems used in different application fields like virtual reality simulation environments, software user interfaces, and digital library systems.A very crucial part of all these systems is the input module which is devoted to recognize the human operator in terms of tracking and/or recognition of human face, arms position, hand gestures, and so on.In this work a software architecture is presented, for the automatic recognition of human arms poses. Our research has been carried on in the robotics framework. A mobile robot that has to find its path to the goal in a partially struc…
HD-RTI: an adaptive multi-light imaging approach for the quality assessment of manufactured surfaces
2021
International audience; Reflectance Transformation Imaging (RTI) is a technique for estimating surface local angular reflectance from a set of stereo-photometric images captured with variable lighting directions. The digitization of this information fully fits into the industry 4.0 approach and makes it possible to characterize the visual properties of a surface. The proposed method, namely HD-RTI, is based on the coupling of RTI and HDR imaging techniques. This coupling is carried out adaptively according to the response at each angle of illumination. The proposed method is applied to five industrial samples which have high local variations of reflectivity because of their heterogeneity of…
Bimanual robot control for surface treatment tasks
2021
This work develops a method to perform surface treatment tasks using a bimanual robotic system, i.e. two robot arms cooperatively performing the task. In particular, one robot arm holds the workpiece while the other robot arm has the treatment tool attached to its end-effector. Moreover, the human user teleoperates all the six coordinates of the former robot arm and two coordinates of the latter robot arm, i.e. the teleoperator can move the treatment tool on the plane given by the workpiece surface. Furthermore, a force sensor attached to the treatment tool is used to automatically attain the desired pressure between the tool and the workpiece and to automatically keep the tool orientation …