This paper presents a part of our work on the development of an embedded stereo vision sensor for mobile robotic platforms. In more detail, it deals with the development and evaluation of stereo matching algorithms. F...
详细信息
ISBN:
(纸本)9783901509582
This paper presents a part of our work on the development of an embedded stereo vision sensor for mobile robotic platforms. In more detail, it deals with the development and evaluation of stereo matching algorithms. For simulation of matching algorithms we have developed a framework that offers rapid prototyping capabilities. It is PC based and used for algorithm simulation. The resulting disparity maps are evaluated using a pixel wise quality formulation. Furthermore, ideas for a taxonomy with embedded implementation possibilities of algorithms as classification criteria are presented To reach the maximum processing power, parallel processing on embedded systems is taken under consideration, too.
Wireless sensor networks (WSNs) typically consist of a large number of densely populated sensor nodes. Due to important advances in integrated circuits and radio technologies, the use of distributed sensor networks is...
详细信息
Wireless sensor networks (WSNs) typically consist of a large number of densely populated sensor nodes. Due to important advances in integrated circuits and radio technologies, the use of distributed sensor networks is becoming increasingly widespread for a variety of applications, e.g. indoor navigation, environmental monitoring, people and object tracking, logistics, industrial diagnostics, quality control, and other manufacturing activities. In many cases, such as in objects tracking, knowing the physical location of network nodes is essential. Locating elements of WSNs is not it trivial task. Manual methods are wearisome and may be inaccurate, especially for large-scale networks. Therefore, many self-locating methods - where nodes cooperate with each other without human involvement - have recently been Studied and implemented. The purpose of this work is to analyse the most significant methods for automatic location of distributed WSNs. The first part of the paper provides a description of the most common criteria used to categorize existing network localization algorithms. A taxonomy is then suggested that may be a useful tool to help evaluate, compare and select such algorithms. Five of the most representative algorithms are explained and discussed in detail in order to identify their strong points and their limitations.
暂无评论