Christophe Ducottet

Discriminative Color Descriptors

Color description is a challenging task because of large variations in RGB values which occur due to scene accidental events, such as shadows, shading, specularities, illuminant color changes, and changes in viewing geometry. Traditionally, this challenge has been addressed by capturing the variations in physics-based models, and deriving invariants for the undesired variations. The drawback of this approach is that sets of distinguishable colors in the original color space are mapped to the same value in the photometric invariant space. This results in a drop of discriminative power of the color description. In this paper we take an information theoretic approach to color description. We cluster color values together based on their discriminative power in a classification problem. The clustering has the explicit objective to minimize the drop of mutual information of the final representation. We show that such a color description automatically learns a certain degree of photometric invariance. We also show that a universal color representation, which is based on other data sets than the one at hand, can obtain competing performance. Experiments show that the proposed descriptor outperforms existing photometric invariants. Furthermore, we show that combined with shape description these color descriptors obtain excellent results on four challenging datasets, namely, PASCAL VOC 2007, Flowers-102, Stanford dogs-120 and Birds-200.

Digital in-line holography: influence of the reconstruction function on the axial profile of a reconstructed particle image

The numerical reconstruction of an in-line digital hologram is a critical point in digital holographic particle image velocimetry. In particular, the shape of the axial profile of the reconstructed particles plays an important role in depth recovery. We show that this profile presents some oscillations when reconstructing by convolution with the Fresnel function. A window can be introduced in the expression of the reconstruction function in order to control these oscillations. The effects of this windowing are discussed and a criterion for the choice of window is given. The method is then illustrated by the processing of a digital hologram of Lycopode particles.

Twin-image noise reduction by phase retrieval in in-line digital holography

In-line digital holography conciles the applicative interest of a simple optical set-up with the speed, low cost and potential of digital reconstruction. We address the twin-image problem that arises in holography due to the lack of phase information in intensity measurements. This problem is of great importance in in-line holography where spatial elimination of the twin-image cannot be carried out as in off-axis holography. Applications in digital holography of particle fields greatly depend on its suppression to reach greater particle concentrations, keeping a sufficient signal to noise ratio in reconstructed images. We describe in this paper methods to improve numerically the reconstructed images by twin-image reduction.

Direct extraction of the mean particle size from a digital hologram

Digital holography, which consists of both acquiring the hologram image in a digital camera and numerically reconstructing the information, offers new and faster ways to make the most of a hologram. We describe a new method to determine the rough size of particles in an in-line hologram. This method relies on a property that is specific to interference patterns in Fresnel holograms: Self-correlation of a hologram provides access to size information. The proposed method is both simple and fast and gives results with acceptable precision. It suppresses all the problems related to the numerical depth of focus when large depth volumes are analyzed.

Scale-adaptive detection and local characterization of edges based on wavelet transform

In this paper, we present an edge detection and characterization method based on wavelet transform. This method relies on a modelization of contours as smoothed singularities of three particular types (transition, peak and line). Using the wavelet transform modulus maxima lines of the edge models, position and descriptive parameters of each edge point can be inferred. Indeed, on the one hand, the proposed algorithm allows to detect and locate edges at a locally adapted scale. On the other hand, it is able to identify the type of each detected edge point and to measure both its amplitude and smoothness degree. The latter parameters represent, respectively, the contrast and the blur level of the edge point. Evaluation of the method is performed on both synthetic and real images. Synthetic data are used to investigate the influence of different factors and the sensitivity to noise, whereas real images allow to highlight the performance and interests of the method. In particular, we point out that the measured smoothness degree provides a cue to recover depth from defocused images or a cue to diffusion measurements in images of cloud structures. Moreover, from an indoor scene, we demonstrate the relevance of type identification for segmentation purposes.

Pattern matching using morphological probing

In this paper, we introduce two new morphological transforms for pattern matching in gray scale images. They rely on a profiling approach and are defined in the context of mathematical morphology. The first transform allows to detect all occurrences of a single pattern in an image, which justifies the name SOMP (single object matching using probing). It is shown to have the properties of a metric and therefore returns a measure of similarity between the search image and the reference pattern. Other properties relative to noise and computation time are highlighted. The second transform MOMP (multiple objects matching using probing) offers the ability to locate multiple patterns simultaneously. It is particularly suited to the detection of objects varying in size and with noisy distortion. Some results are presented for both transforms.

String representations and distances in deep Convolutional Neural Networks for image classification

Recent advances in image classification mostly rely on the use of powerful local features combined with an adapted image representation. Although Convolutional Neural Network (CNN) features learned from ImageNet were shown to be generic and very efficient, they still lack of flexibility to take into account variations in the spatial layout of visual elements. In this paper, we investigate the use of structural representations on top of pretrained CNN features to improve image classification. Images are represented as strings of CNN features. Similarities between such representations are computed using two new edit distance variants adapted to the image classification domain. Our algorithms have been implemented and tested on several challenging datasets, 15Scenes, Caltech101, Pascal VOC 2007 and MIT indoor. The results show that our idea of using structural string representations and distances clearly improves the classification performance over standard approaches based on CNN and SVM with linear kernel, as well as other recognized methods of the literature. HighlightsA structural representation of images on top of CNN features is proposed.Images are represented as strings to integrate spatial relationships.We introduce tailored string edit distances to compare images represented as strings.Experiments show that our structural approach is more powerful than existing ones.It also outperforms state-of-the-art CNN-based classification methods.

Spatial orientations of visual word pairs to improve Bag-of-Visual-Words model.

This paper presents a novel approach to incorporate spatial information in the bag-of-visual-words model for category level and scene classification. In the traditional bag-of-visual-words model, feature vectors are histograms of visual words. This representation is appearance based and does not contain any information regarding the arrangement of the visual words in the 2D image space. In this framework, we present a simple and effi- cient way to infuse spatial information. Particularly, we are interested in explicit global relationships among the spatial positions of visual words. Therefore, we take advantage of the orientation of the segments formed by Pairs of Identical visual Words (PIW). An evenly distributed normalized histogram of angles of PIW is computed. Histograms pro- duced by each word type constitute a powerful description of intra type visual words relationships. Experiments on challenging datasets demonstrate that our method is com- petitive with the concurrent ones. We also show that, our method provides important complementary information to the spatial pyramid matching and can improve the overall performance.

Numerical suppression of the twin image in in-line holography of a volume of micro-objects

We address the twin-image problem that arises in holography due to the lack of phase information in intensity measurements. This problem is of great importance in in-line holography where spatial elimination of the twin image cannot be carried out as in off-axis holography. A unifying description of existing digital suppression methods is given in the light of deconvolution techniques. Holograms of objects spread in 3D cannot be processed through available approaches. We suggest an iterative algorithm and demonstrate its efficacy on both simulated and real data. This method is suitable to enhance the reconstructed images from a digital hologram of small objects.

Tracking granules on the Sun's surface and reconstructing velocity fields I. The CST algorithm

Aims. Determination of horizontal velocity fields on the solar surface is crucial for understanding the dynamics of structures like mesogranulation or supergranulation or simply the distribution of magnetic fields. Methods. We pursue here the development of a method called CST for coherent structure tracking, which determines the horizontal motion of granules in the field of view. Results. We first devise a generalization of Strous method for the segmentation of images and show that when segmentation follows the shape of granules more closely, granule tracking is less effective for large granules because of increased sensitivity to granule fragmentation. We then introduce the multi-resolution analysis on the velocity field, based on Daubechies wavelets, which provides a view of this field on different scales. An algorithm for computing the field derivatives, like the horizontal divergence and the vertical vorticity, is also devised. The effects from the lack of data or from terrestrial atmospheric distortion of the images are also briefly discussed.