A mathematical framework for the solution of statistical inference problems on a class of random sets is proposed. It is based on a new definition of expected pattern. The least-mean-difference estimator (restoration ...
详细信息
A mathematical framework for the solution of statistical inference problems on a class of random sets is proposed. It is based on a new definition of expected pattern. The least-mean-difference estimator (restoration filter) is proved, under certain conditions, to be equivalent to the minimization of the measure of size (area) of the set-difference between the original pattern and the expected pattern of the estimated (restored) pattern. Consequently, it is proved that, under certain conditions, if the estimator (restoration filter) is unbiased, then it is the least mean difference estimator (restoration filter).< >
A hardware and software system for the generation of video patterns that can be used for various vision tests, particularly depth perception, is presented. Among the capabilities of the system is generation of static ...
详细信息
A hardware and software system for the generation of video patterns that can be used for various vision tests, particularly depth perception, is presented. Among the capabilities of the system is generation of static and dynamic random dot stereograms that can be used to test depth perception. The system generates the patterns in real time, which provides the ability of moving the objects in the stereogram. The system is based on the TMS34010 graphics processor, operating at 40 MHz. The processor is interfaced to 512 Kb of video random access memory (VRAM), 256 Kb of dynamic random access memory (DRAM), and 16 Kb of erasable programmable read only memory (EPROM). The system also has an RS-232C serial communication part for communicating with a host computer and a red, green, and blue (RGB) video output for connection to an RGB color video monitor which displays the various video patterns. The software developed produces the various patterns and test experiments. A monitor program was also written to aid in developing any new software needed to produce new patterns or test experiments.< >
Intelligent robotic systems should be capable of performing a range of complex tasks in unstructured and dynamic environments. This chapter discusses a number of important items associated with intelligent robots, rol...
详细信息
Intelligent robotic systems should be capable of performing a range of complex tasks in unstructured and dynamic environments. This chapter discusses a number of important items associated with intelligent robots, role of vision sensing in intelligent robotics, and architecture for the design of an intelligent robot. The chapter also describes the computational task hierarchy that governs a model-based robot vision system, and presents a detailed tutorial on two important components of this hierarchy: image segmentation and matching. The main capabilities that intelligent robotic systems should possess include ability for perception of their environment, planning of their actions, and execution of the planned action. The two important components of a model-based vision system are image segmentation and the matching module. A procedure that allows transformation of 2D image coordinates into 3D world coordinates is also discussed in the chapter. The architecture and control aspects of an integrated sensor-based robotic system consist of six modules: supervisor, task planner, perception, motor, user interface, and knowledge-base. Intelligent, sensor-driven systems pose many challenging basic-research problems from a broad range of scientific and technological disciplines; however, there are many important application areas where such systems would be of significant value and utility, including hazardous environments such as nuclear plants, underwater, and space as well as many industrial-automation tasks. Recent advancements in the VLSI, computer-architectures, sensor-and robot-hardware, and artificial-intelligence fields may aid to the development of robust, practical, and cost-effective intelligent robotic systems for a wide variety of applications.
We present a method for making accurate, optimally localized measurements of the fractal dimension of images modeled as locally fractal Brownian surfaces. Fractal Brownian surfaces are good models for the multiscale a...
Image smoothing and segmentation algorithms are frequently formulated as optimization problems. Linear and nonlinear (reciprocal)resistive networks have solutions characterized by an extremum principle. Thus, appropri...
A description is given of research efforts directed towards the development of an active exploration system performing acquisition, processing, and integration of sensory inputs from vision, point laser range, tactile...
详细信息
A description is given of research efforts directed towards the development of an active exploration system performing acquisition, processing, and integration of sensory inputs from vision, point laser range, tactile, and force/torque sensors. An attempt is made to bring out the rationale that robotic workcell exploration can benefit by utilizing noncontact (e.g., vision, range, and proximity) as well as contact (e.g., tactile and force/torque) sensors while accomplishing tasks such as object identification, object localization, object grasping and object manipulation. The experimental results presented indicate that the rationale utilized in the design of the multisensory robotic active exploration scheme can work successfully.< >
In active laser range finders, the position of the center of gravity of the imaged laserspot is used in triangulation computations for the recovery of range data. The computation of reliable range data thus depends on...
详细信息
The image reconstruction capacity of the actual foveal cone sampling mosaics of an adult monkey and human is investigated through computer simulations. A retinocortical mapping model in which positions are known and d...
We examine the tracking of 3-dimensional targets moving in a complex (e.g. highly textured) visual environment, which makes the application of methods relying on static segmentation and feature correspondence very pro...
详细信息
暂无评论