Rémy Leconge

A New Adaptive Switching Median Filter

A new Adaptive Switching Median (ASWM) filter for removing impulse noise from corrupted images is presented. The originality of ASWM is that no a priori Threshold is needed as in the case of a classical Switching Median filter. Instead, Threshold is computed locally from image pixels intensity values in a sliding window. Results show that ASWM provides better performance in terms of PSNR and MAE than many other median filter variants for random-valued impulse noise. In addition it can preserve more image details in a high noise environment.

Indoor navigation assistance with a Smartphone camera based on vanishing points

Indoor navigation assistance is a highly challenging task that is increasingly needed in various types of applications such as visually impaired guidance, emergency intervention, tourism, etc. Many alternative techniques to GPS have been explored to deal with this challenge like pre-installed sensor networks (Wifi, Ultra Wide Band, Bluetooth, Radio Frequency IDentification etc), inertial sensors or camera. This paper presents an indoor navigation system on Smartphone that was designed taking into consideration low cost, portability and the lightweight of the used algorithm in terms of computation power and storage space. The proposed solution relies on embedded vision. Robust and fast camera orientation (3 dof) is estimated by tracking three orthogonal vanishing points in a video stream acquired with the camera of a free-handled Smartphone. The developed algorithm enables indoor pedestrian localization in two steps: an off-line learning step defines a reference path by selecting key frames along the way using saliency extraction method and computing the camera orientation in these frames. Then, in localization step, an approximate but realistic position of the walker is estimated in real time by comparing the orientation of the camera in the current image and that of reference to assist the pedestrian with navigation guidance. Unlike SLAM, this approach does not require to build 3D mapping of the environment. Online walking direction is given by Smartphone camera which advantageously replaces the compass sensor since it performs very poorly indoors due to electromagnetic noise. Experiments, executed online on Smartphone, that show the feasibility and evaluate the accuracy of the proposed positioning approach for different indoor paths.

Comparison of Image Restoration Methods for Bioluminescence Imaging

Bioluminescence imaging is a recent modality to visualize biological effects, especially for small animals. However, the acquired images are degraded by diffusion and absorption phenomena from the tissue and by the acquisition system itself. In this paper, we use restoration methods to enhance the quality of bioluminescence images. We propose a model for image formation and an experimental determination of the PSF (Point Spread Function). Several restoration methods are compared on test images generated according to the model and on real data. This comparison is insured by using MSE (Mean Square Error) and two other quantitative criteria. Results showed that the statistical methods give more accurate restoration and are well adapted for Bioluminescence Imaging.

Real-time camera orientation estimation based on vanishing point tracking under Manhattan World assumption

Abstract This paper proposes a real-time pipeline for estimating the camera orientation based on vanishing points for indoor navigation assistance on a Smartphone. The orientation of embedded camera relies on the ability to find a reliable triplet of orthogonal vanishing points. The proposed pipeline introduces a novel sampling strategy among finite and infinite vanishing points with a random sample consensus-based line clustering and a tracking along a video sequence to enforce the accuracy and the robustness by extracting the three most pertinent orthogonal directions while preserving a short processing time for real-time application. Experiments on real images and video sequences acquired with a Smartphone show that the proposed strategy for selecting orthogonal vanishing points is pertinent as our algorithm gives better results than the recently published RNS optimal method, in particular for the yaw angle, which is actually essential for the navigation task.

Tracking orthogonal vanishing points in video sequences for a reliable camera orientation in Manhattan World

This paper proposes an algorithm pipeline for estimating the camera orientation based on vanishing points computation targeting pedestrian navigation assistance in Manhattan World. Inspired from some of published methods, the proposed pipeline introduces a novel sampling strategy among finite and infinite vanishing points and a tracking along a video sequence to enforce the robustness by extracting the three most pertinent orthogonal directions while preserving a short processing time for real-time application. Experiments on real images and video sequences show that the proposed heuristic strategy for selecting orthogonal vanishing points is pertinent as our algorithm gives better results than the recently published RNS optimal method [16], in particular for the yaw angle, which is actually essential for navigation task.

Hand gesture recognition using a dedicated geometric descriptor

A high proportion of hospital-acquired diseases are transmitted nowadays during surgery despite existing asepsis preservation measures. These are quite drastic, prohibiting surgeons from interacting directly with non-sterile equipment. Indirect control is presently achieved through an assistant or a nurse. Gesture-based Human-Computer Interfaces constitute a promising approach for giving direct control over such equipment to surgeons. This paper introduces a novel hand descriptor based on measurements extracted from hand contour convex and concave extrema. Using a 9750-picture database created especially for this purpose, it is compared with three state-of-the-art description methods, namely Hu moments, and both SIFT and HOG features. Effects of large amounts of hand rotation are also studied on each rotation axis independently. Obtained results give HOG features as best in recognizing hands from our database, closely followed by the proposed descriptor. Performance comparison when facing rotated hands shows our descriptor as the most robust to rotations, outperforming the other descriptors by a wide margin.

A new detector for switching median filter

In this paper, a new detector for switching median filter is presented. The originality of this approach is that no a priori threshold is to be given. Instead, it is automatically computed from image pixels and based on the weighted mean and the weighted variance in a selected sliding window. The weights are inversely proportional to the grey levels difference between each pixel of the considered window and the mean value of this window. Results show that this new algorithm provides better performances in terms of PSNR and MAE than many other variants of switching median filter. It suppresses noise and preserves details. In addition, psycho visual results are of high quality.

3D reconstruction method of the proximal femur and shape correction

The aim of this work is to present a 3D reconstruction method of the proximal femur shape using contours identification from pairs of 2D X-ray radiographs without any prior acknowledge. 3D personalized model was reconstructed following a processing chain of seven different steps. After localization of the 2D contours on the images and the matching points of these contours, a 3D contour is generated using an algorithm based on a mathematical model. Thus, with a reduced number of pairs of images, we reconstruct a 3D points cloud, which enables obtaining a closed 3D surface. The accuracy of our approach was evaluated by comparing the reconstruction result with the 3D CT-scan reconstruction of cadaveric proximal femur. The estimated error shows that it is possible to rebuild the proximal femur shape from a limited number of radiographs.

Hand posture recognition with multiview descriptors

Preservation of asepsis in operating rooms is essential for limiting the contamination of patients by hospital-acquired infections. Strict rules hinder surgeons from interacting directly with any sterile equipement, requiring the intermediary of an assistant or a nurse. Such indirect control may prove itself clumsy and slow up the performed surgery. Gesture-based Human-Computer Interfaces show a promising alternative to assistants and could help surgeons in taking direct control over sterile equipements in the future without jeopardizing asepsis. This paper presents the experiments we led on hand posture feature selection and the obtained results. State-of-the-art description methods classified in four different categories (i.e. local, semi-local, global and geometric description approaches) have been selected to this end. Their recognition rates when combined with a linear Support Vector Machine classifier are compared while attempting to recognize hand postures issued from an ad-hoc database. For each descriptor, we study the effects of removing the background to simulate a segmentation step and the importance of a correct hand framing in the picture. Obtained results show all descriptors benefit to various extents from the segmentation step. Geometric approaches perform best, followed closely by Dalal et al.s Histogram of Oriented Gradients.