作者: Richard J. Radke
DOI: 10.1007/978-0-387-93808-0_2
关键词:
摘要: Over the past twenty years, computer vision community has made great strides in automatic solution to such problems as camera localization and visual tracking. Many algorithms have been tractable by rapid increases computational speed memory size now available a single computer. However, world of sensor networks poses several challenges direct application traditional algorithms. First, are assumed contain tens hundreds cameras- many more than considered applications. Second, these cameras likely be spread over wide geographical area- much larger typical lab. Third, modest local processors with no ability communicate beyond short range.