Daniel Diep

Texture removal by pixel classification using a rotating filter

In this paper, we present a new method for removing texture in images using a smoothing rotating filter. From this filter, a bank of smoothed images provides pixel signals able to classify a pixel as a texture pixel, a homogenous region pixel or an edge pixel. Then, we introduce a new method for anisotropic diffusion which controls accurately the diffusion near edge and corner points and diffuses isotropically inside textured regions. Several results applied on real images and a comparison with anisotropic diffusion methods show that our model is able to remove the texture and control the diffusion.

Ridges and valleys detection in images using difference of rotating half smoothing filters

In this paper we propose a new ridge/valley detection method in images based on the difference of rotating Gaussian semi filters. The novelty of this approach resides in the mixing of ideas coming both from directional filters and DoG method. We obtain a new ridge/valley anisotropic DoG detector enabling very precise detection of ridge/valley points. Moreover, this detector performs correctly at crest lines even if highly bended, and is precise on junctions. This detector has been tested successfully on various image types presenting difficult problems for classical ridges/valleys detection methods.

RSD-HoG: A New Image Descriptor

In this paper we propose a novel local image descriptor called RSD-HoG. For each pixel in a given support region around a key-point, we extract the rotation signal descriptor(RSD) by spinning a filter made of oriented anisotropic half-gaussian derivative convolution kernel. The obtained signal has extremums at different orientations of the filter. These characteristics are combined with a HoG technique, to obtain a novel descriptor RSD-HoG. The obtained descriptor has rich, discriminative set of local information related to the curvature of the image surface. With these rich set of features, our descriptor finds applications in various branches of computer vision. For evaluation, we have used the standard Oxford data set which has rotation, brightness, illumination, compression and viewpoint changes. Extensive experiments on these images demonstrates that our approach performs better than many state of the art descriptors such as SIFT, GLOH, DAISY and PCA-SIFT.

A new affine invariant method for image matching

This paper describes a new approach in color or grey-scale image matching by points of interest. As many point matching methods, this method is based on two main steps : computation of points and descriptors, followed by a matching process. The points of interest are extracted thanks to the color Harris points detector, they are then described using rotating anisotropic half-gaussian derivative convolution kernels. The descriptors obtained by this filtering stage provide point signatures, robust enough to be recovered in another image even under important color and viewpoint transformations. The matching process uses a cross comparison of point signatures and a voting method to achieve a robust matching. This paper presents the new descriptor defined and the matching process dealing with the data issued from the descriptor.

A Novel Image Descriptor Based on Anisotropic Filtering

In this paper, we present a new image patch descriptor for object detection and image matching. The descriptor is based on the standard HoG pipeline. The descriptor is generated in a novel way, by embedding the response of an oriented anisotropic derivative half Gaussian kernel in the Histogram of Orientation Gradient (HoG) framework. By doing so, we are able to bin more curvature information. As a result, our descriptor performs better than the state of art descriptors such as SIFT, GLOH and DAISY. In addition to this, we repeat the same procedure by replacing the anisotropic derivative half Gaussian kernel with a computationally less complex anisotropic derivative half exponential kernel and achieve similar results. The proposed image descriptors using both the kernels are very robust and shows promising results for variations in brightness, scale, rotation, view point, blur and compression. We have extensively evaluated the effectiveness of the devised method with various challenging image pairs acquired under varying circumstances.

Perceptual curve extraction

In this paper we propose a new perceptual curve detection method in images based on the difference of half rotating Gaussian filters. The novelty of this approach resides in the mixing of ideas coming both from directional filters, perceptual organization and DoG method. We obtain a new anisotropic DoG detector enabling very precise detection of perceptual curve points. Moreover, this detector performs correctly at perceptual curves even if highly bended, and is precise on perceptual junctions. This detector has been tested successfully on various image types presenting real difficult problems for classical detection methods.

Texture Removal Preserving Edges by Diffusion

This article is devoted to a new method for removing texture in images through an image region classification technique using a smoothing rotating filter followed by a diffusion process designed to preserve object contours. This approach lies in associating a descriptor, capable of classifying each pixel as a texture pixel, a homogenous region pixel or an edge pixel, with an anisotropic edge detector serving to define two directions of the edges introduced into an anisotropic diffusion algorithm. Due to the presence of the image region descriptor, the anisotropic diffusion is able to accurately control diffusion near the edges and corner points and moreover remove textured regions. Our results and evaluations based on image segmentation and classical edge detection, which correctly extract objects within the image, compared with anisotropic diffusion methods and nonlinear filters, enable validating our approach.

RSD-DOG: A New Image Descriptor Based on Second Order Derivatives

This paper introduces the new and powerful image patch descriptor based on second order image statistics/derivatives. Here, the image patch is treated as a 3D surface with intensity being the 3rd dimension. The considered 3D surface has a rich set of second order features/statistics such as ridges, valleys, cliffs and so on, that can be easily captured by using the difference of rotating semi Gaussian filters. The originality of this method is based on successfully combining the response of the directional filters with that of the Difference of Gaussian DOG approach. The obtained descriptor shows a good discriminative power when dealing with the variations in illumination, scale, rotation, blur, viewpoint and compression. The experiments on image matching, demonstrates the advantage of the obtained descriptor when compared to its first order counterparts such as SIFT, DAISY, GLOH, GIST and LIDRIC.