Edwige E. Pissaloux

On the Evolution of Parallel Computers Dedicated to Image Processing through Examples of Some French Computers

Abstract This paper proposes an overview of the evolution of the architecture of some parallel computers dedicated to image processing. A formal definition of a dedicated computer is proposed and some requirements of different levels of image processing/vision in terms of their algorithmic structures are pointed out. Examples of few French computers show hardware structures useful for efficient (real-time and on-board) implementations of image processing/vision tasks.

A New Framework for Cognitive Mobility of Visually Impaired Users in Using Tactile Device

This paper proposes a novel framework for a better understanding of human cognitive locomotion and its interaction with the new tactile technologies that assist mobility and the acquisition of spatial knowledge in the absence of sight [case of visually impaired people (VIP), seniors, etc.]. Unlike the existing mobility models, the proposed framework encompasses four elements: 1) walking with obstacle avoidance, 2) orientation, 3) spatial awareness, and 4) actual physical displacement, and is based on the concept of the tactile gist, a refreshable (dynamic) egocentric tactile representation of the perceived scene. The relevance of the tactile gist concept for achieving the assistance of human mobility has been evaluated with three basic original spatial experiments (homing with obstacle avoidance, localization of a passage and its size estimation, and navigation in an apartment) using a purposely designed perception-movement platform and a touch stimulating tangible wearable refreshable device. The collected results show that the tactile gist might be considered as a scene representation and should be integrated into an assistive device for sightless people. The proposed approach may impact the design of mobility aids for VIP as it reinforces their natural mobility skills and may lead to new mobility strategies of VIP in (un-)known environment.

A qualitative image reconstruction from an axial image sequence

This paper presents a method to process axial monocular image sequences for mobile robot obstacle detection. We do not aim to achieve a complete scene reconstruction, but only to evaluate the time to collision and surface orientation useful for robot obstacle avoidance. Using a planar facet representation we first calculate formally the velocity field generated by the camera motion. The apparent deformations, in conjunction with a projective model, are then used in order to evaluate the scene apparent movement with a wide angle camera. In practice, we process separately the tangential and radial components of the apparent velocity vectors, using the epipolar constraint. Noise resistance is improved by integration using the Greens and Stokes theorems which provide a link with surface moments. Experimental results on synthesis and real images of indoor scenes are given, and their validity is discussed Potential applications include visual navigation, obstacle detection, visual servoing, and object recognition.

Multispectral segmentation and textural feature extraction

This paper focuses on the design of segmentation based on textural features extraction. This is an example of a transposition between biological and visual phenomena used in order to characterize natural image understanding. This is also an illustration of a more general appraoch to IP knowledge representation based on a methodology dedicated to the formalization of concrete and abstract models for image processing applications. It proposes an ontology which includes conceptual specifications borrowed from mathematics and physical and biological axiomatics which give concrete and more natural sense to our IP models. This provides a set of elementary definitions which can be used for the expression of concrete models such as image segmentation or pattern detection. In the case of texture we would like to formalize grey level behavior through processings based on multiple window analysis (spectral and morphological criteria: grey level and compacity). In this framework, the evolutionary models studied are issued from biological modeling of migration and mutation. Our illustration is relevant to multispectral segmentation where the homogeneity criterion has been modelized by a grey level evolution function based on exponentiation.

Model of Cognitive Mobility for Visually Impaired and its Experimental Validation

This paper reports the results of three experiments on tactile perception, spatial representation, and tactile display design for assistive mobility devices for blind, partially sighted, and deaf blind people. The results indicate the potentials of tactile displays for supporting environmental exploration and mobility. Voluntary test subjects showed reasonably good ability to determine the direction of motion of an arrow, with best recognition rates in the up and right directions. They showed reasonably good ability to use a tactile display to detect and avoid obstacles after a very short learning period and more limited ability to learn and remember an environmental representation using information from a tactile display and walking through the environment without specific instructions.

On Spatial Cognition and Mobility Strategies

People acquire spatial knowledge by physically experiencing the environment through locomotion. Spatial knowledge generally emerges from the interactions between the specific types of sensory data and the cognitive strategies involved in locomotion. This chapter reviews the fundamental concepts and evidences of spatial cognition. First, the interplay between visual data and mobility is addressed. The importance of landmarks and visual cues is highlighted for implementing both simple and complex navigation strategies such as path integration (PI), landmark-based, and geometry-based, which are the main cognitive mechanisms for spatial learning. Some neural evidences of spatial cognition are given to underline the brain mechanisms involved in cognitive mapping. A review of relevant literature models on cognitive mapping is also presented to better understand how spatial representations are formed at the level of the brain. The chapter provides some insights on how to adapt all these concepts for mobility and spatial knowledge assistance of people with visual impairments.

A vision-application adaptable computer concept and its implementation in FreeTIV computer

This paper introduces a concept of an adaptable (or polymorph) architecture named Free-TIV computer (FREE architecture dedicated to image processing and vision). This computer architecture for vision applications has the capability to adapt its internal structure to the run-time requirements for vision algorithms. Such a concept allows one to reuse the same available computation resources in different applications, and thus to develop a fast and reliable, dedicated or not, computer at a low cost.An implementation of the FreeTIV is proposed as well. FreeTIV is a heterogeneous run-time reconfigurable parallel computer encompassing several processing modules. Different module interconnections are provided by a message passing router called RouTIV (a router dedicated to image processing and vision). RouTIV adapts FreeTIV module interconnections to the data movements of vision algorithms. Some details related to the implementation of RouTIV with Xilinx Virtex 300 are provided. An example of an image matching algorithm running on FreeTIV is provided.

Toward a system for automatic FPGA-based reconfigurable vision system design

This paper proposes a new methodology for rapid virtual system design and prototyping guided by a set of system basic architectures. This methodology includes the automatic hardware/software partitioning of a considered applications. The partitioning is performed under user-specified constraints (temporal performance, real time, system volume, ...). The aimed plat-form for system realization is a PC board with DSP and FPGA. The proposed methodology will be applied for design of a generic router node for an autonomous robot image processing board.

Massively parallel implementation of image processing algorithms

This paper focuses on the problem of the massively parallel implementation of image processing algorithms. In previous theoretical studies the parallel software requirements to implement image processing algorithms were pointed out. A test algorithm, which is representative enough of the requirements for edge and region segmentations was chosen. Our goal here is to detail its implementation. The proposed test algorithm was implemented with the data programming model, on the connection machine CM5 in C*, which is an extension of the C programming language. The crux points of the parallel implementation are underlined. Edge point detection requires only parallel operations, and regular communications. Conversely, region extraction and edge chaining require irregular communications, therefore for a better efficiency, in both cases, the original algorithms were modified. These studies are in relation with the problem of finding tools and methods to compare and to evaluate parallel architectures. One of the two proposed algorithms is deduced of this one.

Towards Haptic Surface Devices with Force Feedback for Visually Impaired People.

This paper presents a new haptic surface tablet that can provide force feedback to the user. Force feedback means that the device can react to the user’s movements and apply a force against or in-line with these movements, according to the tactile properties of a displayed image. The device consists of a frame attached to a tactile tablet that generates a force feedback to user’s finger when exploring the surface, providing haptic informations about the displayed image. The experimental results suggest the relevance of this tablet as an assistive device for visually impaired people in perceiving and understanding the content of a displayed image. Several potential applications are briefly presented.