system’s components, supporting theoretical ideas that are provided by a Gaussian pyramid, whereas by experimental results. Section 2 introduces a our framework also utilizes intermediate levels. Also, frame-differencing technique used to satisfy the real-time execution would require that a high- goal of visually detecting objects of interest. Special performance image processor (like the VLSI chip effort is made to develop algorithms that allow the developed by van der Wal and Burt'') be available to system to function with a consideration for real-time compute the pyramids. constraints. Section 3 addresses additional chal- The frame-differencing system by Dinstein'' lenges that an automated detection and tracking provides an interesting alternative to the figure system must often consider, including analysis of segmentation approach described as a part of our visual detection results. Section 4 presents the framework. Dinstein uses four projections to identify various visual calculations that are used in order to the centers of multiple moving objects in a fashion derive the dtsplacements of objects or significant that is only slightly less robust than our method. features within objects. Through these equations, the However, his method locates objects by majority vote system is able to follow the location of objects as they (much like Hough transforms); as a result, the move. Section 5 describes the hardware modules that method does not provide statistics required by are required to operate the visual detection and other portions of our framework (e.g. it cannot tracking framework. Section 6 demonstrates the determine the geometric characteristics of an framework’s use in robotic visual serx oing. Addi- identified object). tional issues for the integration of a detection module Considering application-specific detection, Allen et 机器视觉维修系统英文文献和中文翻译(4):http://www.youerw.com/fanyi/lunwen_80168.html