Frédéric Morain-Nicolier

A topology preserving non-rigid registration algorithm with integration shape knowledge to segment brain subcortical structures from MRI images

A new non-rigid registration method combining image intensity and a priori shape knowledge of the objects in the image is proposed. This method, based on optical flow theory, uses a topology correction strategy to prevent topological changes of the deformed objects and the a priori shape knowledge to keep the object shapes during the deformation process. Advantages of the method over classical intensity based non-rigid registration are that it can improve the registration precision with the a priori knowledge and allows to segment objects at the same time, especially efficient in the case of segmenting adjacent objects having similar intensities. The proposed algorithm is applied to segment brain subcortical structures from 15 real brain MRI images and evaluated by comparing with ground truths. The obtained results show the efficiency and robustness of our method.

Ornamental Letters Image Classification Using Local Dissimilarity Maps

This article describes a new method for ancient books ornamental letters segmentation and recognition. The pur-pose of our work is to automatically determine the letterrepresented in an ornamental letter image. Our process isdivided in two parts: a segmentation step of the ornamentalletter is followed by a recognition step. The segmentationprocess uses multiresolution analysis to filter backgrounddecorations followed by a binarisation step and a morpho-logic reconstruction of the expected letter. The recogni-tion process use the previously obtained reconstruction andcompares it with capital letters images used as a dictionaryof shapes with the Local Dissimilarity Map (LDM) distance.

Retrieval of the Ornaments from the Hand-Press Period: An Overview

This paper deals with the topic of the retrieval of document images focused on a specific application: the ornaments of the Hand-Press period. It presents an overview as a result of the work and the discussions undertaken by a workgroup on this subject. The paper starts by giving a general view about digital libraries of ornaments and associated retrieval problematics. Two main issues are underlined:content based image retrieval (CBIR) and image difference visualization. Several contributions are summarized,commented and compared. Conclusions and open problems arising from this overview are twofold: 1. contributions on CBIR miss scale-invariant methods and dont provide significative evaluation results. 2. robust registration is the open problem for visual comparison.

Denoising of dynamic PET images using a multi-scale transform and non-local means filter

Abstract The quantification of positron emission tomography (PET) images requires a time activity curve (TAC) to provide an accurate estimation of kinetic parameters. However, the low signals to noise ratio (SNR), the important level of noise, and the low spatial resolution of PET image make the extraction of the TAC a challenging task. In this study, we present a new method based on multi-scale and non-local means method (MNLM) to reduce noise in dynamic PET sequences of small animal heart. MNLM filter takes into account the temporal correlation between images in the dynamic measurement and benefits from the complementary properties of both the Shearlet transform and the wavelet transform to provide best reduction. The method was tested on dynamic digital mouse phantom and a preclinical rat study (nu202f=u202f6). Based on a comparative study with three major algorithms reviewed on the state of the art, the data analysis proved the significance of the MNLM filter. In simulated data, the major finding of the study showed that at the highest noise level (7.68%), the model gave the best result (Chi-squareu202f=u202f4.06). Furthermore, it presented a notable gain in terms of PSNR and SSIM plot. In real data, the MNLM showed a better result in the computation of the contrast metric with a value of 27.04u202f∓u202f12.1 and the highest SNR with a value of 74.38u202f∓u202f9.2. This approach proved a better potential and could be considered as a valuable candidate to reduce noise in clinical system.

Multi-scale and Non Local Mean based filter for Positron Emission Tomography imaging denoising

Dynamic Positron Emission Tomography (PET) is a functional imaging modality which provides information about tracer kinetic in a specific target. In the last three decades, the [18F]-fluorodeoxyglucose ([18F]-FDG) tracer has been widely used by many institutions to measure the local myocardium metabolic rate for glucose. The analysis of the dynamic measurements requires, often, parameters estimation in which the PET data is noisy. In this paper, we propose a systematic methodology to reduce noise in PET data based on the combination of an extension of Non Local Means algorithm and the Discrete Curvelet Transform. The methodology was applied to a small animal model study of the heart, where both the input function and the tissue tracer concentrations at each time were derived from de-noised images. Experimental results revealed a significant improvement in SNR and the spatial distribution of the tracer.

Gray Level Local Dissimilarity Map and Global Dissimilarity Index for Quality of Medical Images

Abstract In order to evaluate performance quality of coding techniques, it is needed to have a good global index and a local index allowing the localisation of the distortions. In this study, a local dissimilarity map is presented for gray-level images. Its application to the comparison of a compressed image and its reference allows an excellent visual detection of the distortions. A global dissimilarity index is computed from the local dissimilarity map. These new measures are compared to the structural similarity index (SSIM). The results of the global measure are as good as the SSIM. The results of the local measure are quite superior to the SSIM computed in a local window. We claim these good results come from the consistency of the proposed index. It is more consistent to compute a global measure from a local one, than a local measure from a global one.

Non-rigid registration based segmentation of brain subcortical structures using a priori knowledge

Segmentation of the brain internal structures is an important and a challenging task due to their complex shapes, partial volume effects, low contrasts and anatomical variability between subjects. In this paper we propose a new non-rigid registration method that automatically segments the deep brain internal structures from brain MRI images. An atlas of the structures is used as a priori knowledge, which is modeled as a shape representation. By integrating the shape knowledge into a classical intensity based non-rigid registration algorithm, the proposed segmentation method allows to ameliorate the results in the case of low contrast on the boundaries of the structures. The shape model is based on distance representation obtained from the atlas. The segmentation of brain subcortical structures is performed on real MRI images and the obtained results are very encouraging.

Segmentation of Brain Internal Structures Automatically Using Non-rigid Registration with Simultaneous Intensity and Geometric Match

Segmentation of the brain internal structures is an important and a challenging task due to their small size, partial volume effects, and anatomical variability. In this paper we propose a method that segments automatically the deep brain internal structures from brain MRI images. It uses a combination of local affine transformation and optical flow based non-rigid registration, which has the advantages of modifying the larger geometric deformation and intensity differences simultaneously. Meanwhile the residual subtle differences decrease due to the high degree of freedom. Both simulated data and real data are used to validate the proposed method and the results are encouraging. It can be concluded that the image gray level of the corresponding structures plays an important role in registration based segmentation using intensity metric.

Face Spoofing Detection for Smartphones using a 3D Reconstruction and the Motion Sensors.

Face recognition system is proven to be vulnerable to face spoofing attack. Many approaches have been proposed in the literature to resolve this vulnerability. This paper proposes a novel method dedicated to mobile systems. The approach asks users to capture a video by moving the device around their face. Thanks to a 3D reconstruction process, the shape of the object is estimated from the video. By evaluating this 3D shape, we can rapidly eliminate attacks in which a photo of a legitimate face is used. Then, the camera’s poses estimated from the 3D reconstruction is used to be compared to the data captured from the device’s motion sensors. Experimental results on a real database show the efficiency of the proposed approach.

Face spoofing attack detection based on the behavior of noises

This paper aims to study the problem of spoofing attack detection for facial recognition systems. Real faces and falsified faces present in front of a security system (phones camera in our case) have differences of micro-textures on their surface, which are exploited to discriminate face spoofing images. Our method exploits the statistic behavior of the distribution of noises local variances, which performs differently between images of real faces and the fake ones. We test our method on two databases constructed in our laboratory. We used SVM for classification method. Experimental results show that the proposed method has an encouraging performance.