The calibration is discussed in two main parts: for the photon Detector Assembly and for the Camera Module containing optical and other parts. Some of the problems raised can be pursued with model calculations and sim...
详细信息
The calibration is discussed in two main parts: for the photon Detector Assembly and for the Camera Module containing optical and other parts. Some of the problems raised can be pursued with model calculations and simulations, but others will not be solved until real calibration data are available.
A clustering algorithm is employed for the compression of the very high rate data generated. The information transmitted to ground is for each line, an address map containing the cluster number each pixel belongs to, ...
详细信息
A clustering algorithm is employed for the compression of the very high rate data generated. The information transmitted to ground is for each line, an address map containing the cluster number each pixel belongs to, and a dictionary containing the cluster centers. A compression ratio of 8. 7 is achieved when the image consists of 6 spectral channels.
Standard techniques for enhancement, image-restoration, and the removal of camera-induced distortion developed and used previously at the imageprocessing Laboratory in the Planetary Program were employed. In addition...
详细信息
Standard techniques for enhancement, image-restoration, and the removal of camera-induced distortion developed and used previously at the imageprocessing Laboratory in the Planetary Program were employed. In addition, new digital processing techniques have been formulated to explore atmospheric dynamics, spectral variations, and the morphology of Jupiter, Saturn, and their satellites.
After a brief description of the instrument and its operating modes, its scientific objectives are summarized. The basic data processing required to remove instrument signature effects from the images is then discusse...
详细信息
After a brief description of the instrument and its operating modes, its scientific objectives are summarized. The basic data processing required to remove instrument signature effects from the images is then discussed. More subtle errors in the data are then presented and the techniques for their removal identified. In general, these secondary errors require interaction between the observer and the computer for their complete elimination. The data processing techniques envisioned for extracting astronimical information from the images are then addressed.
The conference proceedings contains 22 papers, of which 10 are indexed separately. The topics discussed are the following: Earth Resources Technology Satellite (ERTS); ERTS/space telescope; planetary imaging; PI/coded...
详细信息
The conference proceedings contains 22 papers, of which 10 are indexed separately. The topics discussed are the following: Earth Resources Technology Satellite (ERTS); ERTS/space telescope; planetary imaging; PI/coded mask telescopes; imageprocessing systems and techniques; and astronomical analysis.
This paper describes a modular imageprocessing software-hardware system (IPS) developed at McGill University for imageprocessing and patternrecognition research in medicine and biology. The IPS Software runs on a P...
详细信息
This paper describes a modular imageprocessing software-hardware system (IPS) developed at McGill University for imageprocessing and patternrecognition research in medicine and biology. The IPS Software runs on a PDP11 computer system equipped with several image acquisition and display devices. It is in regular use in a variety of research applications including: the development of an automated prescreening system for cervical cytology specimens, studies of cerebral blood flow, and studies of metabolism in bone.
An outline of some of the activities in computer-based image understanding in the Department of computer Science at the University of British Columbia is given. The introduction focuses on paradigms for image analysis...
详细信息
An outline of some of the activities in computer-based image understanding in the Department of computer Science at the University of British Columbia is given. The introduction focuses on paradigms for image analysis and discusses the development of image understanding within artificial intelligence. Techniques from that area have been combined with modifications of traditional patternrecognition techniques to enhance the spectral, spatial and semantic sensitivity of Landsat image segmentation programs. Two programs, one using interpretation-guided region merging and the other using image hierarchies, are briefly presented. Further work on understanding sketch maps and the like shows how to make explicit and then exploit, the cartographic and geographic semantics of simple maps, and, moreover, contributes directly to the new computational theory of perception.
The current status of the development of an on-board navigational update system for aircraft is discussed. The system is unique in that the update is accomplished by the application of imageprocessing and pattern rec...
详细信息
The current status of the development of an on-board navigational update system for aircraft is discussed. The system is unique in that the update is accomplished by the application of imageprocessing and patternrecognition techniques to provide for the detection of known terrestrial features as they are displayed on a navigator's screen.
Problems of efficient measurement selection and classification are investigated in the context of discrimination between mixtures of parametric densities. Bounds on the Bayes error rate are obtained, and a recursive a...
详细信息
Problems of efficient measurement selection and classification are investigated in the context of discrimination between mixtures of parametric densities. Bounds on the Bayes error rate are obtained, and a recursive algorithm is proposed for comparing discriminant functions and error bounds.
A k-nearest neighbor rule based upon the notion of empirical distance is discussed that is Bayes risk consistent for all possible distributions of the observations and has the desired invariance property.
A k-nearest neighbor rule based upon the notion of empirical distance is discussed that is Bayes risk consistent for all possible distributions of the observations and has the desired invariance property.
暂无评论