Sequential Detection and Classification in 3D Computer Vision

序列观测统计行为的最快检测和分类问题是一个经典问题，在工程、经济和流行病学中有着广泛的应用。 在当今快速发展的技术中，新的应用领域不断涌现。特别是，城市场景的自动三维图像重建和分类是一个复杂的问题，仍然挑战计算机科学家。 传统上，它是通过使用激光扫描仪采集数据来处理的，激光扫描仪可以产生高分辨率的图像，但速度可能非常慢。因此，必须将激光扫描仅集中到感兴趣的区域，这导致对感兴趣区域的快速决策。这可以节省大量的时间和成本，同时仍然可以生成高分辨率的3D图像。 该项目的目标是开发和实现实时算法，用于处理和分析3D激光测距数据。通过巧妙地选择创新的测量模型，减少了数据的高维性质。然后，通过在线参数和非参数分类和检测技术来处理相互依赖的观测流。最后，新的统计模型被用来捕捉城市场景中的障碍物。这提供了一个系统的处理问题的快速和有效的三维图像分类使用高分辨率的激光数据。目前的建议预计将为传统的最快检测和分类技术开发和建立一系列新的可能性。该项目有望将经典序列统计的应用扩展到3D计算机视觉领域，从而为创建具有明确共同目标的计算机科学家和统计学家的新的协同跨学科研究和教育团队开辟道路。结合这两个社区的专业知识，预计将导致更复杂的技术和软件，用于采集和处理来自激光扫描仪的3D数据。该项目为本科生创造了一个刺激的研究环境，激励他们寻求数学和计算机科学跨学科前沿的高级研究。它还提供了一个框架，通过开发跨学科课程，创新布鲁克林学院这一少数群体服务机构的课程。

The problem of quickest detection and classification in the statistical behavior of sequential observations is a classical one, with numerous applications in engineering, economics and epidemiology. In today's fast-growing technologies new areas of applications constantly emerge. In particular, the automatic 3D image reconstruction and classification of urban scenes is a problem whose complexity still challenges computer scientists. It has traditionally been treated through the acquisition of data using laser-scanners, which produce high-resolution images, but can be very slow. It is thus essential to concentrate laser scanning only to the areas of interest, which leads to fast decision-making about areas of interest. This can save significant time and cost, while still producing high-resolution 3D images. The goal of this project is to develop and implement real-time algorithms for processing and analyzing 3D laser range data. The high-dimensional nature of the data is reduced by a clever innovative selection of a measurement model. Interdependent streams of observations are then processed by on-line parametric and non-parametric classification and detection techniques. And finally, new statistical models are used to capture obstacles in urban scenes. This provides a systematic treatment of the problems of fast and efficient 3D image classification using high-resolution laser data. The current proposal is expected to develop and establish a new line of possibilities for the traditional quickest detection and classification techniques. This project promises to expand the applications of classical sequential statistics to the area of 3D Computer vision, thus carving the road for the creation of new synergistic interdisciplinary research and education teams of computer scientists and statisticians with well-defined common goals. Combining the expertise of these two communities is expected to lead to more sophisticated technology and software for the acquisition and processing of 3D data that comes from laser scanners. This project creates a stimulating research environment for undergraduate students, motivating them to seek advanced studies on the interdisciplinary frontier of mathematics and computer science. It also provides a framework for innovating the curriculum at Brooklyn College, a minority-serving institution, through the development of interdisciplinary courses.