IPOLIPOLhttp://www.ipol.im/feed/IPOL Preprints — Latest public preprints from IPOL.ikiwiki2019-07-21T22:29:17ZCloud Detection by Disparity Phase Analysis for Pushbroom Satellite Imagershttp://www.ipol.im/pub/pre/271/Tristan Dagobert,
Rafael Grompone von Gioi,
Carlo de Franchis,
Jean-Michel Morel2019-07-21T22:29:17Z2019-07-21T22:27:49Z
This paper proposes a cloud detector for earth observation images obtained by a push-broom satellite imager. This particular technology induces an inter-channel acquisition delay leading to a parallax effect for the clouds. We propose a method exploiting this characteristic thanks to the analysis of the visible inter-bands disparity maps. Several other features discriminating clouds are also defined and all are merged to build a robust a contrario statistical decision. Experiments applied on scenes acquired by various push-broom satellites such that Sentinel-2, Rapideye and Worldview-2 show the effectiveness of the proposed method. In particular, we demonstrate 97% and 98% success rates for cloud and non cloud classification for Sentinel-2 images.
The Gradient Product Transform: An Image Filter for Symmetry Detectionhttp://www.ipol.im/pub/pre/270/Christoph Dalitz,
Jens Wilberg2019-07-21T22:14:15Z2019-07-21T22:14:15Z
The Gradient Product Transform (GPT) is an image filter that converts a gray scale image into a float image, such that points representing a point reflection symmetry center obtain a high score. Beside the symmetry score, it also yields an estimator for the size of the symmetry region around each point. Apart from describing the GPT, the article also explains its application for two use cases: detection of objects with a point reflection or C2m rotational symmetry, and the extraction of blood vessel skeletons from medical images. For the detection of symmetric objects, a score normalization procedure is suggested that allows to choose a fixed threshold for score values representing actual symmetries.
Cross-comparison of the Performance of Sequential Summed Area Table and Box Filter Algorithms with respect to C/C++ Compilershttp://www.ipol.im/pub/pre/268/Ali Ozturk,
Ibrahim Cayiroglu2019-07-21T22:03:34Z2019-07-21T22:03:34Z
Summed area table algorithm has been used to accelerate some computer vision and signal processing algorithms. In this study, the performance of the sequential summed area table algorithms and box filter algorithm with and without summed area table algorithm is examined by taking account of effect of C/C++ compilers. Three variants of sequential summed area table algorithm are included into the study. Loop invariant code motion and loop unrolling optimization techniques are applied to one of them. The performance of GNU Compiler Collection (GCC) and Intel C/C++ Compiler (ICC) on both Windows and Ubuntu and Visual C++ (CL) compiler on Windows by using the summed area table and box filter algorithms are compared. Result of the study reveals that Intel C/C++ Compiler (ICC) perform best on Ubuntu with respect to others for sequential summed area table and box filter algorithm. The performance of summed area table calculation which utilizes Viola-Jones Equation by using a scalar accumulator outperforms other summed area table algorithms.
A Histogram Based Approach for Contrast Enhancementhttp://www.ipol.im/pub/pre/264/Zardoua Yassir,
Boulaala Mohammed2019-07-21T21:48:06Z2019-07-21T21:48:06Z
In this paper, we present a detailed histogram based approach to enhance contrast of images. We aim to stretch the histogram by establishing a mapping function composed of multiple stretching segments (MSS). The slope of each segment is mainly controlled by the probability density of pixels on the range it covers. The use of a maximum number of segments led us to the formula of Global Histogram Equalization (GHE), which is obtained by a different manner. The difference in the approach of GHE and MSS when using the maximum number of segments is explained in detail.
How to Reduce Anomaly Detection in Images to Anomaly Detection in Noisehttp://www.ipol.im/pub/pre/263/Thibaud Ehret,
Axel Davy,
Mauricio Delbracio,
Jean-Michel Morel2019-07-21T14:56:26Z2019-07-21T14:55:42Z
Anomaly detectors address the difficult problem of detecting automatically exceptions in a background image, that can be as diverse as a fabric or a mammography. Detection methods have been proposed by the thousands because each problem requires a different background model. By analyzing the existing approaches, we show that the problem can be reduced to detecting anomalies in residual images (extracted from the target image) in which noise and anomalies prevail. Hence, the general and impossible background modeling problem is replaced by a simple noise model, and allows the calculation of rigorous detection thresholds. Our approach is therefore unsupervised and works on arbitrary images. The residual images can favorably be computed on dense features of neural networks. Our detector is powered by the a contrario detection theory, which avoids over-detection by fixing detection thresholds taking into account the multiple tests.
Optimization of Image B-Spline Interpolation for GPU Architectureshttp://www.ipol.im/pub/pre/257/Thibaud Briand,
Axel Davy2019-07-20T15:06:20Z2019-07-20T14:57:01Z
Interpolation is a vital piece of many image processing pipelines. In this document we present how to optimize the implementation of image B-spline interpolation for GPU architectures. The implementation is based on the one proposed by Briand and Monasse in 2018 and works for orders up to 11. The two main optimizations consist in: (1) transposing the B-spline coefficients before the prefiltering of the rows and (2) dividing columns into subregions in order to use more threads. We assess the impact of the floating point precision and of using high B-spline orders.
Visibility Detector for Time Series of Spectrally Limited Satellite Imagershttp://www.ipol.im/pub/pre/245/Tristan Dagobert,
Jean-Michel Morel,
Carlo de Franchis2019-07-20T15:20:39Z2019-07-19T16:25:37Z
This article addresses the problem of estimating scene visibility in time
series of satellite images. We are especially focused on satellites with few
spectral bands but with high revisit frequency. Our approach exploits the
redundancy of information acquired during these revisits. It is based on
an unsupervised algorithm that tracks local ground textures across time
and detects ruptures caused by opaque clouds, haze, cirrus and shadows.
Experiments have been carried out on 18 Planet times series representing
various locations. These time series come with hand-made labeled ground
truth that we make available to the scientific community. We compare the
results obtained with those of the PlanetScope algorithm and demonstrate
the effectiveness of the proposed method: success rates of 94% and 84% are
reached for the visible and occulted regions classification.
Matching of Weakly-Localized Features Under Different Geometric Modelshttp://www.ipol.im/pub/pre/247/Erez Farhan2019-07-19T16:10:59Z2019-07-19T16:10:59Z
Matching corresponding local patches between images is a fundamental building block in many
computer-vision algorithms, reducing the high-dimensional challenge of recovering geometric
relations between images to a series of relatively simple and independent tasks. This approach
is geometrically very flexible and has clear computational advantages over more convoluted
global solutions. But it also has two major practical shortcomings: 1) Sparsity: the need to rely
on high-quality repeatable features for matching drives current local methods to discard low-
textured image locations and leave them unanalysed; 2) Reliability: the limited spatial context
in which those methods work often does not contain enough information for achieving reliable
matches. In this work, we target a major blind spot of local feature matching: ill-textured
locations. We observe that while classic methods avoided using poorly localized features (e.g.
edges) as matching candidates, due to their low reliability, these features contain highly valuable
information for image registration. We show how, given the appropriate geometric context,
reliable matches can be produced from these features, contributing to a better coverage of the
scene. We present a statistically attractive framework for encoding the uncertainty that stems
from using weakly localized matches into a coupled geometric estimation and match extraction
process. We examine the practical application of the proposed framework to the problems of
guided matching and affine region expansion and show significant improvement over preceding
methods.
The Orthographic Projection Model for Pose Calibration of Long Focal Imageshttp://www.ipol.im/pub/pre/248/Laura F. Julià,
Pascal Monasse,
Marc Pierrot-Deseilligny2019-07-19T15:57:39Z2019-07-19T15:57:39Z
Most stereovision and Structure from Motion (SfM) methods rely on the pinhole camera model
based on perspective projection. From this hypothesis the fundamental matrix and the epipolar
constraints are derived, which are the milestones of pose estimation. In this article we present
a method based on the matrix factorization due to Tomasi and Kanade that relies on a simpler
camera model, resulting in orthographic projection. This method can be used for the pose
estimation of perspective cameras in configurations where other methods fail, in particular,
when using cameras with long focal length lenses. We show this projection is an approximation
of the pinhole camera model when the camera is far away from the scene. The performance of
our implementation of this pose estimation method is compared to that given by the
perspective-based methods for several configurations using both synthetic and real data. We show through
some examples and experiments that the accuracy achieved and the robustness of this method
make it worth considering in any SfM procedure.
A Contrario Detection of Faces with a Short Cascade of Classifiershttp://www.ipol.im/pub/pre/272/Jose-Luis Lisani,
Silvia Ramis2019-07-19T12:33:22Z2019-07-19T12:32:10Z
The a contrario framework has been succesfully used for the detection of lines, contours and
other meaningful structures in digital images. In this paper we describe the implementation of
an algorithm for face detection published in 2017 by Lisani et al. which applies the a contrario
approach to the computation of the detection thresholds of a classical cascade of classifiers. The
result is a very short cascade which improves the detection rates of a classical (and longer) one
at a much lower computational cost.
Extraction of the Level Lines of a Bilinear Imagehttp://www.ipol.im/pub/pre/269/Pascal Monasse2019-06-29T14:03:29Z2019-06-29T14:00:24Z
We detail precisely an algorithm for the extraction of the level lines of a bilinear image, which is a continuous function interpolating bilinearly a discrete image. If we discard the levels of the discrete image, where topological difficulties arise, a level line is a concatenation of branches of hyperbolas. The algorithm tracks these branches and provides a sampling of the level lines in the form of closed polygons. If the level line contains a saddle point, the hyperbola degenerates to orthogonal segments, where an arbitrary but consistent choice is adopted for the tracking at the bifurcation. In any case, the extracted polygons are disjoint and enclose a bounded region. This allows to order the level lines in an enclosure tree hierarchy, which may be used for a variety of filters. Recovering this tree is a simple post-processing of the extraction algorithm.
A Data Set for the Study of Human Locomotion with Inertial Measurements Unitshttp://www.ipol.im/pub/pre/265/Charles Truong,
Rémi Barrois-Müller,
Thomas Moreau,
Clément Provost,
Aliénor Vienne-Jumeau,
Albane Moreau,
Pierre-Paul Vidal,
Nicolas Vayatis,
Stéphane Buffat,
Alain Yelnik,
Damien Ricard,
Laurent Oudre2019-06-23T16:31:36Z2019-06-23T16:31:36Z
This article thoroughly describes a data set of 1020 multivariate gait signals collected with two inertial measurement units, from 230 subjects undergoing a fixed protocol: standing still, walking 10 m, turning around, walking back and stopping. In total, 8.5 h of gait time series are distributed. The measured population was composed of healthy subjects as well as patients with neurological or orthopedic disorders. An outstanding feature of this data set is the amount of signal metadata that are provided. In particular, the start and end time stamps of more than 40,000 footsteps are available, as well as a number of contextual information about each trial. This exact data set was used in Oudre, et al. (2018) to design and evaluate a step detection procedure.
A Parallel, O (n) Algorithm for an Unbiased, Thin Watershedhttp://www.ipol.im/pub/pre/215/Théodore Chabardès,
Petr Dokládal,
Matthieu Faessel,
Michel Bilodeau2019-05-09T10:30:32Z2019-05-09T10:30:32Z
The watershed transform is a powerful tool for morphological segmentation. Most common
implementations of this method involve a strict hierarchy on gray tones in processing the pixels
composing an image. This hierarchical dependency complexifies the efficient use of modern
computational architectures. This paper introduces a new way of computing the watershed
transform that alleviates the sequential nature of hierarchical queue propagation. It is shown
that this method can directly relate to the hierarchical flooding. Simultaneous and disorderly
growth can now be used to maximize performances on modern architectures. Higher speed is
reached, bigger data volume can be processed. Experimental results show increased performances
regarding execution speed and memory consumption.
An Implementation of the Mean Shift Algorithmhttp://www.ipol.im/pub/pre/255/Damir Demirović2019-05-09T10:23:49Z2019-05-09T10:23:49Z
In this paper we present an implementation and analysis of Mean shift algorithm in C++
language. Mean shift is a general non-parametric mode finding/clustering procedure widely
used in image processing and analysis and computer vision techniques such as image denosing,
image segmentation, motion tracking etc.
Watervoxelshttp://www.ipol.im/pub/pre/250/Pierre Cettour-Janet,
Clément Cazorla,
Vaia Machairas,
Quentin Delannoy,
Nathalie Bednarek,
François Rousseau,
Etienne Décencière,
Nicolas Passat2019-05-31T07:49:18Z2019-05-09T10:14:40Z
In this article, we present the n-dimensional version of the waterpixels, namely the
watervoxels. Waterpixels constitute a simple, yet efficient alternative to standard superpixel paradigms,
initially developed in the field of computer vision for reducing the space cost of input images
without altering the accuracy of further image processing / analysis procedures. Waterpixels
were initially proposed in a 2-dimensional version. Their extension to 3-dimensions—and more
generally n-dimensions—is however possible, in particular in the Cartesian grid. Indeed, water-
pixels mainly rely on a seeded watershed transformation applied on a saliency map defined as
the linear combination of a gradient map and a distance map. We propose a description of the
algorithmics of watervoxels in n-dimensional Cartesian grids. We also discuss its parameters
and its time cost. A source code for 2- and 3-dimensional versions of watervoxels is provided,
such as a 2-dimensional demonstrator. This article can be seen as the companion of the article
“Waterpixels”, published in 2015 in IEEE Transactions on Image Processing.
An Analysis and Implementation of Natural Image Enhancement Methodhttp://www.ipol.im/pub/pre/252/Dang Thanh Trung,
Shaohua Chenr,
Azeddine Beghdadi,
Nguyen thi Quynh Hoa2019-03-16T21:56:11Z2019-03-16T21:56:11Z
Digital image enhancement is to improve the appearance of images to human viewers. This is one of extremely difficult issues in image processing. The NECI algorithm (Natural Enhancement of Color Image) is a robust and effective approach for color image enhancement. The core of this algorithm is Retinex model, a simulation model of human color vision. A fully framework for enhancement process is composed of four main steps: global tone mapping, Retinex-based local contrast enhancement, histogram rescaling, and texture enhancement.
The experimental results on different types of natural images show that the proposed algorithm not only preserves the ambience of image but also avoid side effects such as light condition changes, color temperature alteration, or additional artifacts, etc which lead to unnaturally sharpened images or dramatic white balance changes. In the output color images, no additional light sources are added to the scene, or no halo effect and blocking effect are amplified due to over-enhancement. Moreover, all parameters are image-dependent so that the process requires no parameter tuning.
Comparison of Optical Flow Methods under Stereomatching with Short Baselineshttp://www.ipol.im/pub/pre/217/Tristan Dagobert,
Nelson Monzón,
Javier Sánchez2019-01-21T12:49:29Z2017-10-16T12:24:59Z
This article studies the effectiveness of optical flow methods applied to short
baseline image pairs under different noise levels. New metrics have been
developed to analyze the results because the usual metrics are inadequate in a
subpixel context. We have used the implementation of some standard optical flow
methods adapted to the stereo problem. Our experiments show that the Brox
et al. method produces the least errors, with a 60% success rate and a
relative precision at 1/100th of a pixel. On the other hand, our comparison
shows that a discontinuity preserving method, derived from Brox et al.,
also provides competitive results at the same time that it yields disparities
with more details and correct contours.
Joint Large-Scale Motion Estimation and Image Reconstructionhttp://www.ipol.im/pub/pre/193/Hendrik Dirks2018-07-22T11:40:36Z2016-11-24T13:13:55Z
This article describes the implementation of the joint motion estimation and image reconstruction framework presented by Burger, Dirks and Schönlieb and extends this framework to large-scale motion between consecutive image frames. The variational framework uses displacements between consecutive frames based on the optical
flow approach to improve the image reconstruction quality on the one hand and the motion estimation quality on the other. The energy functional consists of a
data-fidelity term with a general operator that connects the input sequence to the solution, it has a total variation term for the image sequence and is connected to the underlying flow using an optical flow term. Additional spatial regularity for the flow is modeled by a total variation regularizer for both components of the flow. The numerical minimization is performed in an alternating manner using
primal-dual techniques. The resulting schemes are presented as pseudo-code together with a short numerical evaluation.
Joint TV-L1 Optical Flow and Occlusion Estimationhttp://www.ipol.im/pub/pre/118/Juan Francisco Garamendi Bragado,
Coloma Ballester,
Lluís Garrido,
Vanel Lazcano,
Vicent Caselles2018-07-22T11:40:36Z2015-02-05T12:37:40Z
This document describes an implementation of the energy functional minimization proposed by Ballester, Garrido, Lazcano and Caselles for joint optical flow and occlusion estimation. The method is based on the TV-L1 approach introduced Zach, Pock and Bischof in 2007 but with the particularity of detecting occlusions. The energy functional is composed by a regularization term (over the optical flow and the occlusion fields) using the total variation, a data term using the L1 norm, and a term, which is based on the divergence of the flow, for dealing with the occlusions.