A new type of range sensing method which is based on the principle of triangulation has been developed. A mirror tunnel is placed between an observation lens and an object to be measured;a light beam is projected thro...
详细信息
ISBN:
(纸本)0818605456
A new type of range sensing method which is based on the principle of triangulation has been developed. A mirror tunnel is placed between an observation lens and an object to be measured;a light beam is projected through the observation lens on to the object;the projected light spot is imaged by the observation lens on the image plane after reflection with the mirror;a position of the light spot on the object can be determined according to the triangulation. The width of an optical system can be reduced to remarkably smaller than the effective base line of the triangulation, and it may be suitable to miniaturize the size of range sensor such as a proximity sensor.
The proceedings contains 375 papers. The papers are grouped under the following headings: motion;biomedical applications;edge detection;similarity and distance;texture;curves;feature selection;imageprocessing;speech;...
详细信息
ISBN:
(纸本)0818605456
The proceedings contains 375 papers. The papers are grouped under the following headings: motion;biomedical applications;edge detection;similarity and distance;texture;curves;feature selection;imageprocessing;speech;mobile robotics;quadtrees;implementation and complexity;classification algorithms;segmentation;relaxation and interference;shape transforms;efficient data structures;oriental character recognition;remotesensing;applications;3-D vision;computational geometry;linear features;human perception;VLSI implementation;imageprocessing and shape;character recognition;3-D robotics;systems;patternrecognition;data structures and coding;image understanding;shape matching;hardware implementation;signatures and shape;drawing analysis;fuzzy methods;classification;microelectronic inspection;and syntactic methods.
作者：
Mauer, E.Research Inst for Information
Processing & Pattern Recognition Ettlingen West Ger Research Inst for Information Processing & Pattern Recognition Ettlingen West Ger
The basic methods for image segmentation and object classification, which have been developed until now, work separately on the diverse image properties. In order to accumulate the advantages of each method a spatial ...
详细信息
The basic methods for image segmentation and object classification, which have been developed until now, work separately on the diverse image properties. In order to accumulate the advantages of each method a spatial selective application, depending on the data properties, of the different methods is necessary. To optimize the combination of methods, the different decision thresholds of the process are automatically adapted to each other and to the actual data properties. Thus, a high degree of adaptive automation has been attained. An implemented automatic system for multispectral image classification without a priori knowledge is described.
作者：
Shapiro, Linda G.Virginia Polytechnic Inst &
State Univ Dep of Computer Science Blacksburg VA USA Virginia Polytechnic Inst & State Univ Dep of Computer Science Blacksburg VA USA
The backtracking search procedures used to solve consistent labeling problems have exponential complexity in the number of units. Even the addition of relaxation operators cannot reduce the complexity for some problem...
详细信息
ISBN:
(纸本)0818605456
The backtracking search procedures used to solve consistent labeling problems have exponential complexity in the number of units. Even the addition of relaxation operators cannot reduce the complexity for some problems. There are, however, some classes of problems that can be solved by simpler procedures. One such class is the set of consistent labeling problems having the separation property. In this paper, the author defines this class of problems (which arise in remotesensing), gives a divide-and-conquer algorithm for their solution, and discusses the complexity of the algorithm. Present results indicate expected linear behavior.
作者：
Akey, Mark L.Mitchell, O.R.Purdue Univ
Sch of Electrical Engineering West Lafaytte IN USA Purdue Univ Sch of Electrical Engineering West Lafaytte IN USA
This paper deals with the detection and precision location of ground targets in the form of synthetic crosses on real terrain data. Location of crosses using Fourier descriptors can be achieved to within 0. 07 pixels....
详细信息
ISBN:
(纸本)0818605456
This paper deals with the detection and precision location of ground targets in the form of synthetic crosses on real terrain data. Location of crosses using Fourier descriptors can be achieved to within 0. 07 pixels. This method is used to monitor the geometric distortion caused by compressing the data with the cosine transform. Results show geometric shifts up to 0. 5 pixels may occur after compression. Mean and median filters are applied to synthetic test data. These filters show a minimal amount of geometric shift in the presence of noise. The mean filter reduces location error to 0. 025 pixels. A new circular two-dimensional median filter is introduced and is shown to instill less geometric distortion than the conventional rectangular median filter.
Combining information that can be derived from intensity range images allows for the direct computation of motion parameters. Intensity images which can be obtained rapidly are used as a basis for guiding the extracti...
详细信息
ISBN:
(纸本)0818605456
Combining information that can be derived from intensity range images allows for the direct computation of motion parameters. Intensity images which can be obtained rapidly are used as a basis for guiding the extraction of a small set of correspondence points whose three-dimensional coordinates can be determined by a ranging device. Model centered motion parameters are computed by a least squares approach. Object centered motion parameters are derived by computing differences between model centered parameters for successive views. Results are presented for synthetic range and intensity images.
This conference proceedings contains 171 papers. Among the topics covered are: 1-D digital filters and design methods;2-D and M-D digital filters and design methods;digital transformations;spectral estimation;adaptive...
详细信息
ISBN:
(纸本)0444875832
This conference proceedings contains 171 papers. Among the topics covered are: 1-D digital filters and design methods;2-D and M-D digital filters and design methods;digital transformations;spectral estimation;adaptive processing;implementation techniques and architectures;special devices and dedicated realizations;VLSI processors;digital signal processing techniques;speech processing;digital imageprocessing;digital signal processing in communications;radar applications;remotesensing;biomedical signal and imageprocessing;and patternrecognition and robotics.
These issues contain 14 conference papers. The topics covered are: target detection and subpixel location;terrain elevation and surface feature data analysis;knowledge-based aerial photo interpretation;aerial photogra...
详细信息
These issues contain 14 conference papers. The topics covered are: target detection and subpixel location;terrain elevation and surface feature data analysis;knowledge-based aerial photo interpretation;aerial photograph segmentation;image sequence comparison;and image coding. 8 of these papers have been abstracted previously for COMPENDEX.
This conference proceedings contains 47 papers. Among the areas covered are: sensor based manufacturing;vision systems;sensor guided welding;3-dimensional sensing;sensory control;non-vision sensing;knowledge-based sen...
详细信息
ISBN:
(纸本)0903608758
This conference proceedings contains 47 papers. Among the areas covered are: sensor based manufacturing;vision systems;sensor guided welding;3-dimensional sensing;sensory control;non-vision sensing;knowledge-based sensory systems;advanced vision techniques. The topics discussed include: vision driven picking;iconic imageprocessing;eddy current sensors;machine vision algorithms;object recognition;real time path control of a robot;tactile sensing arrays;magnetostrictive skin for robots.
Accurate and reliable detection of unique objects is an important component of an image understanding system. The objects are considered to have a unique pattern and should be recognized based upon their own character...
详细信息
ISBN:
(纸本)0892525207
Accurate and reliable detection of unique objects is an important component of an image understanding system. The objects are considered to have a unique pattern and should be recognized based upon their own characteristics. An object detection approach applicable to the high resolution aerial images is developed in the paper. The methodologies incorporate features such as minimal sensitivity to the backgrounds on which objects may appear, ability to detect objects appearing in arbitrary orientations, and use of a unified set of operators to analyze various levels of details. Specifically the gray level co-occurrence matrix is used to provide the early vision operators. Results of several experiments show the promise for the proposed methods.
暂无评论