François Pasteau

Vision-based adaptive assistance and haptic guidance for safe wheelchair corridor following

We devise a low-cost vision-based assistance scheme for wheelchair corridor navigation.Progressive assistance is activated if the wheelchair is in danger, as the user drives manually in a corridor.Natural image features that can be extracted robustly in real-time is employed in order to design the assistive controller.Optimal force feedback, also designed from visual information, is provided using a haptic joystick in order to increase intuitiveness of the system.Experimental results show the feasibility of the system for Commercialization. In case of motor impairments, steering a wheelchair can become a hazardous task. Joystick jerks induced by uncontrolled motions may lead to wall collisions when a user steers a wheelchair along a corridor. This work introduces a low-cost assistive and guidance system for indoor corridor navigation in a wheelchair, which uses purely visual information, and which is capable of providing automatic trajectory correction and haptic guidance in order to avoid wall collisions. A visual servoing approach to autonomous corridor following serves as the backbone to this system. The algorithm employs natural image features which can be robustly extracted in real time. This algorithm is then fused with manual joystick input from the user so that progressive assistance and trajectory correction can be activated as soon as the user is in danger of collision. A force feedback in conjunction with the assistance is provided on the joystick in order to guide the user out of his dangerous trajectory. This ensures intuitive guidance and minimal interference from the trajectory correction system. In addition to being a low-cost approach, it can be seen that the proposed solution does not require an a-priori environment model. Experiments on a robotised wheelchair equipped with a monocular camera prove the capability of the system to adaptively guide and assist a user navigating in a corridor.

Quality constraint and rate-distortion optimization for predictive image coders

Next generations of image and video coding methods should of course be efficient in terms of compression, but also propose advanced functionalities. Among these functionalities such as scalability, lossy and lossless coding, data protection, Rate Distortion Optimization (RDO) and Rate Control (RC) are key issues. RDO aims at optimizing compression performances, while RC mechanism enables to exactly compress at a given rate. A less common functionality than RC, but certainly more helpful, is Quality Control (QC): the constraint is here given by the quality. In this paper, we introduce a joint solution for RDO and QC applied to a still image codec called Locally Adaptive Resolution (LAR), providing scalability both in resolution and SNR and based on a multi-resolution structure. The technique does not require any additional encoding pass. It relies on a modeling and estimation of the prediction errors obtained in an early work. First, quality constraint is applied and propagated through the whole resolution levels called pyramid. Then, the quantization parameters are deduced considering inter and intra pyramid level relationships. Results show that performances of the proposed method are very close to an exhaustive search solution.

Fast pseudo-semantic segmentation for joint region-based hierarchical and multiresolution representation

In this paper, we present a new scalable segmentation algorithm called JHMS (Joint Hierarchical and Multiresolution Segmentation) that is characterized by region-based hierarchy and resolution scalability. Most of the proposed algorithms either apply a multiresolution segmentation or a hierarchical segmentation. The proposed approach combines both multiresolution and hierarchical segmentation processes. Indeed, the image is considered as a set of images at different levels of resolution, where at each level a hierarchical segmentation is performed. Multiresolution implies that a segmentation of a given level is reused in further segmentation processes operated at next levels so that to insure contour consistency between different resolutions. Each level of resolution provides a Region Adjacency Graph (RAG) that describes the neighborhood relationships between regions within a given level of the multiresolution representation. Region label consistency is preserved thanks to a dedicated projection algorithm based on inter-level relationships. Moreover, a preprocess based on a quadtree partitioning reduces the amount of input data thus leading to a lower overall complexity of the segmentation framework. Experiments show that we obtain effective results when compared to the state of the art together with a lower complexity.

Locally Adaptive Resolution (LAR) codec

The JPEG committee has initiated a study of potential technologies dedicated to future generation image compression systems. The idea is to design a new norm of image compression, named JPEG AIC (Advanced Image Coding), together with advanced evaluation methodologies, closely matching to human vision system characteristics. JPEG AIC thus aimed at defining a complete coding system able to address advanced functionalities such as lossy to lossless compression, scalability (spatial, temporal, depth, quality, complexity, component, granularity...), robustness, embed-ability, content description for image handling at object level... The chosen compression method would have to fit perceptual metrics defined by the JPEG community within the JPEG AIC project. In this context, we propose the Locally Adaptive Resolution (LAR) codec as a contribution to the relative call for technologies, tending to fit all of previous functionalities. This method is a coding solution that simultaneously proposes a relevant representation of the image. This property is exploited through various complementary coding schemes in order to design a highly scalable encoder. The LAR method has been initially introduced for lossy image coding. This efficient image compression solution relies on a content-based system driven by a specific quadtree representation, based on the assumption that an image can be represented as layers of basic information and local texture. Multiresolution versions of this codec have shown their efficiency, from low bit rates up to lossless compressed images. An original hierarchical self-extracting region representation has also been elaborated: a segmentation process is realized at both coder and decoder, leading to a free segmentation map. This later can be further exploited for color region encoding, image handling at region level. Moreover, the inherent structure of the LAR codec can be used for advanced functionalities such as content securization purposes. In particular, dedicated Unequal Error Protection systems have been produced and tested for transmission over the Internet or wireless channels. Hierarchical selective encryption techniques have been adapted to our coding scheme. Data hiding system based on the LAR multiresolution description allows efficient content protection. Thanks to the modularity of our coding scheme, complexity can be adjusted to address various embedded systems. For example, basic version of the LAR coder has been implemented onto FPGA platform while respecting real-time constraints. Pyramidal LAR solution and hierarchical segmentation process have also been prototyped on DSPs heterogeneous architectures. This chapter first introduces JPEG AIC scope and details associated requirements. Then we develop the technical features, of the LAR system, and show the originality of the proposed scheme, both in terms of functionalities and services. In particular, we show that the LAR coder remains efficient for natural images, medical images, and art images.