Small format, medium resolution CCD cameras are at present widely used for industrial metrology applications because they are readily available and relatively inexpensive. The calibration of CCD cameras is necessary in order to characterize the geometry of the sensors and lenses. In cases where a static or slowly moving object is to be imaged, frame capture mode is most often used to maximize the resolution of the CCD sensor. In cases where the object to be imaged is in rapid motion, field mode capture is frequently adopted to avoid discontinuities caused by the interlaced scanning process used by the sensors. A strategy for the computation of calibration parameters for field modes, based on a frame mode calibration, is proposed. The strategy is then verified by testing using real data from test range calibrations of three different types of CCD camera. The results of these calibration tests are described, with particular emphasis on the changes required to adjust between frame and field capture modes.

Automated digital photogrammetric systems are considered to be passive three-dimensional vision systems since they obtain object coordinates from only the information contained in intensity images. Active 3-D vision systems, such as laser scanners and structured light systems obtain the object coordinates from external information such as scanning angle, time of flight, or shape of projected patterns. Passive systems provide high accuracy on well defined features, such as targets and edges however, unmarked surfaces are hard to measure. These systems may also be difficult to automate in unstructured environments since they are highly affected by the ambient light. Active systems provide their own illumination and the features to be measured so they can easily measure surfaces in most environments. However, they have difficulties with varying surface finish or sharp discontinuities such as edges. Therefore each type of sensor is more suited for a specific type of objects and features, and they are often complementary. This paper compares the measurement accuracy, on various type of features, of some technologically-different 3-D vision systems: photogrammetry-based (passive) systems, a laser scanning system (active), and a range sensor using a mask with two apertures and structured light (active).

The description of the capabilities of Tu-134 and An-2 planes and Mi-6 helicopter testbeds for carrying out remote sensing tasks in the range of 0.4 - 14 micrometer is given. Main features of testbeds are: receipt of synchronous multispectral images with the use of photo-, TV, infrared and laser scanning equipment which is partly installed on gyrostabilized platforms, with registration of data from all optical-electronic devices and navigation equipment into a digital common frame on a high efficient tape recorder; realization of convergent photosurvey and ground objects TV observation in all the upper hemisphere with a possibility of manual or automated (using the correlation algorithm) tracking of selected objects; perfect navigational support, including GPS and GLONASS receivers, inertial system and others and autonomous recording of navigational data.

A multispectral scanner is designed for acquisition of the multispectral spatially matched quantitative information on both reflectance and emittance of on-ground objects. Mechanical scanning is used. Six spectral channels are operable in the range of 0.4 - 12.5 micrometer. All the channels are provided with the removable sets including four optical filters for each channel. Switching or change of the filters can be performed by the remote control unit. The channels are designed to match spectral ranges that are similar to those which can be found in space sensors for TM-84, ETM, SPOT, NOAA. The digitized field of view is 51.2 degrees. The instantaneous field of view for each channel is 5 minutes. The precision of optical matching for all the channels is no worse than 1/3 of pixel. The selectable scan rates are 18, 36, and 72 scans per second. In the channels of the range 0.43 - 0.9 micrometer the multispectral scanner ensures the values of signal/noise ratios 80 - 200; within the range of 1.5 - 2.5 micrometer signal/noise ratios are no less than 100, and within the range of 8 - 12.5 micrometer no less than 25.

The photo compact disk is on its way to becoming a very important tool for every photogrammetrist involved in digital close range applications. The maximum resolution of 2048 by 3072 pixel referring to a 24 by 36 mm negative image, respectively a resolution of 4096 by 6144 with the professional version (Pro Photo CD), leads to the expectation of a high quality standard. The presented paper deals with its data compression method, the analysis of accuracy and geometrical stability of the transferred photo-CD images as a tool for low-cost and easy analog-digital conversion of images. Because the special method of data compression it hits exactly the needs of digital photogrammetry concerning an efficient and fast data storage.

After a historical retrospective the flow of data behind the Leica/Helava systems shows the different steps including scanning process (DSW100, DSW200) combined with aerotriangulation, the transfer of data to the digital photogrammetric workstation (DPW), and the generation of final products. The workstation process flow is explained in detail with DTM and Orthophotos generation.

The use of thermovision systems for the remote sensing of surface temperatures has been conventional for a long time. In recent years serious progress has been achieved in the reduction of instrumental errors of IR receivers and increase of quality of thermal images visualization and registration. The smaller progress is observed in questions of reduction of methodical measuring errors, in a sense of the determination by thermovision methods of the true (thermodynamic) temperature of a radiating surface, its factor of radiation, and the determination of a ratio of own and reflected radiation. It is expected that the essential progress can be achieved by creation of thermovision systems, carrying out registration of objects in several spectral subranges of ranges 3-5 or 8-14 micrometer. A similar result can be achieved by use of single-channel devices with a set of spectral filters. Such an approach provides, in essence, the registration of multispectral thermal images, further mathematical processing of which (or processing in a real time scale), permits us to make approaches to solving the above mentioned problems. Thus, the problem of determination of a spectrum of radiation for each point of the image on the basis of results of multispectral measurements becomes the central one. As a criterion the most optimum accuracy and complete time of measurement (for single-channel systems) or amount of channels (for multichannel systems) should be used. Naturally, the term 'accuracy' must be determined. The offered technique was being created during scientific cooperation between the Institute of Radio-electronics of the Russian Academy of Sciences (IRE RAN), producing thermovision systems, and the State Research Institute of Aviation Systems (GosNIIAS) which makes software for such devices and realizes the multispectral processing.

An on-board system is designed for interrogatory of sensors according to the preset program as well as for normalization, analog-to-digital converting, packing into the frames of the appropriate format and matching with a high-capacity digital tape recorder. Non-synchronous analog and digital units for images acquisition, navigation instruments, and other facilities operate as sensors. System controls the data collection process, forms the data flows with their timing, appropriately logs the data into the on-board tape recorder, generates the operator required signals for the system's main units operational direction. The modular structure allows the configuration to be enlarged and developed according to the feature of the problems being solved.

The revolutionary development of electronic computers from the mid-60s onward, allowed the development of powerful programs for the computation of aerotriangulations (AT). Since the end of the 80s the further explosion in computer-speed and storage-capacity from mega to gigabyte allowed the next step into digital photogrammetry and also to digital aerotriangulation (DAT). The HELAVA DSW 100 opens the possibilities to handle AT with digital data, thus moving away from the actual photograph to scanned digital information, originating from aerial photography or other imagery. It consists of a high-precision scanner, with two CCD sensor cameras and linear encoders for very precise positioning. The actual measurement is done by automatic picture-correlation with the DCCS Software module. The selection of tie and other points can be done on paperprints, no point-transfer-device is needed. The measured picture coordinates are further processed using some existing AT-programs, such as PAT-B ALBANY or in our case BLUH. So far more than 2000 models have been processed. The triangulation data have been used for consecutive work. We could not detect any discrepancies in these consecutive works. For some 40 models, we could compare DAT results with those of a conventional AT, for which an analytical stereoplotter and pugged carry-over-points were used. The results of the DAT were more accurate by a factor of 0.7. Care has to be taken in point-selection, that no disturbing picture-detail complicates automatic correlation. This may be overcome by manual correlation but with inferior accuracy. DAT is certainly a very powerful tool and applicable to a wide range of projects.

With the development of higher-resolution scanners, faster image-handling capabilities, and higher-resolution screens, digital photogrammetric workstations promise to rival conventional analytical plotters in functionality, i.e. in the degree of automation in data capture and processing, and in accuracy. The availability of high quality digital image data and inexpensive high capacity fast mass storage offers the capability to perform accurate semi- automatic or automatic triangulation of digital aerial photo blocks on digital photogrammetric workstations instead of analytical plotters. In this paper, we present our investigations and results on two photogrammetric triangulation blocks, the OEEPE (European Organisation for Experimental Photogrammetric Research) test block (scale 1;4'000) and a Swiss test block (scale 1:12'000) using digitized images. Twenty-eight images of the OEEPE test block were scanned on the Zeiss/Intergraph PS1 and the digital images were delivered with a resolution of 15 micrometer and 30 micrometer, while 20 images of the Swiss test block were scanned on the Desktop Publishing Scanner Agfa Horizon with a resolution of 42 micrometer and on the PS1 with 15 micrometer. Measurements in the digital images were performed on the commercial Digital photogrammetric Station Leica/Helava DPW770 and with basic hard- and software components of the Digital Photogrammetric Station DIPS II, an experimental system of the Institute of Geodesy and Photogrammetry, ETH Zurich. As a reference, the analog images of both photogrammetric test blocks were measured at analytical plotters. On DIPS II measurements of fiducial marks, signalized and natural tie points were performed by least squares template and image matching, while on DPW770 all points were measured by the cross correlation technique. The observations were adjusted in a self-calibrating bundle adjustment. The comparisons between these results and the experiences with the functionality of the commercial and the experimental system are presented.

The software system PHOTOMOD for processing, analysis, and visualization of aerospace and customary pictures is considered. It runs on an IBM-compatible personal computer under WINDOWS 3.1. The main modules and algorithms of the program are descried. Fast and accurate algorithms of digital terrain model creation are suggested. The stereoscopic mode of visualization and results processing are implemented in the system. Test results are presented.

A digital stereophotogrammetric system based on PC is being developed. The system uses standard IBM PC-AT/386-/486 as a processing unit. The system is capable to fulfill processes of stereo model's building and terrain reconstruction using aero and space photographs. The peculiarity of this system is its possibility to process the large digital images exceeding disk memory capacities. Initial images must be previously decomposed on the fragments by use of the Workstation computer. The fragments are accompanied by some extra information. The survey image is created using pyramid image processing.

The number of automatically executed photogrammetrical processes is constantly growing in connection with the development of theory, methods, and technical means of image processing. Advisability of improving these processes is evident especially for those which are, by any indexes, at a disadvantage in relation to non-automatic processes. For example, productivity and precision of automatic stereo measurements are on par with non-automatic, but reliability of stereo measurements are lower. The considerable possibilities of improving the quality of automatic stereo measurements may be found in optimization of the stereo signals identification process: filtration of each stereo signal from noises, taking the information from the stereo image areas being identified, defining and removing the stereo signals discordance. Optimization of each of the processes may demand a special theory and method, not only for the process as a whole, but also for the constituting operations. For instance, while taking information, the following parameters should be optimized: trajectory and moving direction of scanning elements, shape and size of stereo images being identified, number of scanning elements per each part of the area being identified (dispersion of search expenses). Let us tell more about dispersion of search expenses. In existing stereo automats rasters are used, which provide uniform taking of information from image areas being identified, which is not judiciously in the case of its random dispersion, because it leads inevitably to declining of the identification quality. It is impossible to define a judicious strategy of the search beforehand, because of the absence of apriori data about the image areas being investigated. The method has developed, using the theory of random processes management, which optimizes search expenses. In the course of developing the method three rules are defined.

A central projection has been taken as a basic classical method of photogrammetry. A homocentral bundle of projecting rays and planes, which are going through the same point, is the fundamental constructive base in this theory. Terrain points are reflecting into an image by projecting rays, and straight lines -- by projecting planes. Principles of collinearity of a pair of vectors of corresponding points of an image and the terrain and complanearity corresponding straight lines and an image, which are situated in the same plane and intersecting, are going from this theory.

The Modular Optoelectronic Multispectral/Stereo Scanner (MOMS-02) is a newly designed sensor that provides multispectral coverage in 4 wavebands including the visible and near- infrared range. It is also equipped with a three line along-track stereo device recording fore/aft and high resolution nadir panchromatic data. Data acquired during the STS-55/D2 have been investigated with regard to their technical performance and their spectral significance. Different algorithms are used for preprocessing of raw data and merge of panchromatic and multispectral information. With some limitations, radiometry, entropy, SNR and LSF performances are comparable to operational sensors. Coefficients for a transfer function from DN-values to spectral radiance based on inter-sensor calibration to TM are presented. Results of variance/covariance analyses indicate a 10 - 15% improved spectral decorrelation for MOMS-02 bands as compared to those of operational sensors with broader band design.

This paper deals with the design and development of a fast moving surveying system, which meets the demands of highest accuracy asked in geodesy and surveying for determining 3D- positions. The reader is introduced into real time capturing of GPS-, IMU- and CCD-camera outputsignals and into post-mission data processing for automatic digital information acquisition with the kinematic surveying system (KiSS). The absolute position of the moving vehicle within the global coordinate system (WGS84) is obtained by combining data from a GPS receiver and an inertial measuring unit (strapdown IMU). For compensating the errors of both sensors, additional sensing devices like odometer and barometer are utilized. Further improvement in position estimation is achieved by stereo photogrammetric measurements of known environmental objects, the so called landmarks. A digital stereo vision-system creates successive series of high resolution grayscale images, while a SVHS color video system records a continuous image sequence of the traveled road and the nearby surroundings. Depending on the demands of the application, selected environmental objects can be extracted, classified, and photogrammetrically positioned. THe parameters and the images of the acquired objects are indexed to the 3D-trajectory and saved in a GIS-database.

The cost-efficiency of photogrammetric mapping by means of spaceborne stereophotographs heavily depends on the geometric quality of the image. Geometric properties are usually inhomogeneous over the image area and very often only a part of the satellite image is used for practical cartography. In many cases digital image processing under large magnifications is required to make full use of the information capacity of spaceborne photographs with high ground resolution. The present paper deals with the geometric investigation of spaceborne multispectral images obtained by the Russian MK-4 film camera, which is reputed as being one of the best present-day operational satellite camera systems. Since no original photos were at our disposal, the investigations were performed using image copies of the first generation in analog and -- after precise digitizing -- in digital form. Geometric and resolvometric properties of MK-4 imagery taken in different spectral bands were investigated by means of analog and digital methods under large magnifications, that is 16x, 30x, and 90x. Fragmentary image processing was performed to reduce the influence of different geometric distortions and to improve the accuracy of digital stereoplotting and basic large-scale mapping in several areas of the Austrian Alps. Planimetric and elevation accuracies achieved with MK-4 image copies were very satisfactory and sufficient for digital image-based mapping at scales of 1:50,000 and even 1:25,000.

Space based measurements concerning chemistry and physics in the lower atmosphere very often are discussed due to discrepancies found in ground verification by countermeasurements from planes. The manned balloon offers an ideal platform for Lagrange experiments and control measurements staying in defined air masses and heights for longer times. This enables comparing measurements when the launch site is in the field of view of the remote-sensors. The Richard Assmann Launch Site for hydrogen balloons at Ottobrunn, operated since 1992 by Fordergemeinschaft 'Wissenschaftliche Ballonfahrt' e.V. (Fg WB) together with the Daimler-Benz Aerospace A.G. (Dasa), Allows countermeasurements close to conurbanization and agricultural areas a few miles from the German Remote Sensing Center Oberpfaffenhofen. Until now measurements have been executed concerning: (1) Ozone concentration above and below the boundary layer and ozone vertical profiles. (2) Concentration and transport-mechanisms of pesticides above boundary layer. (3) UV background radiation measurements for rocket detection. (4) Optical density of atmosphere in different heights. The Richard Assmann launch site is open for national and international activities. Platform and pilots are available from Fg WB together with technical assistance from Dasa.

In this paper, we introduce an extension to 2-D case of Mallat's algorithm called 'Matching Pursuit'. This algorithm decomposes the image into a linear expansion of waveforms that belong to a redundant dictionary of functions. Using Gabor's space-frequency atoms and a test on the residual image norm, an image compression rate of 31.59 is also reached and a segmentation approach could be realized.

This paper contains information about the geographic information system (GIS)-hardware- software-human-computer complex, which provides fee, handling, displaying and circulating spatial data, data integration and knowledge about terrain for their effective use at decision scientific and applied geographical tasks, relating to analyzing, modeling, forecasting and control of the ambient environment. Geoinformation system contains data about spatial entities, combined in a collection of covers and makes information models of data domain, operation set, which define system functionality and realize GIS technologies software, apparatus and information. GIS prognostic (indicator and estimate) covers display natural phenomena, potentially contain forecast capabilities of their evolution and play the key role in the decision support system. Space photos or remote sensing data, which contain information on natural environment and allow us to allot particular entities and their border, are base data source for special covers in GIS. Below we reveal one of the thematic maps generation methods for a decision support system through the use of instrument environment ERDAS Imagine (ERDAS Inc., Atlanta, Georgia), for environment handling and remote sensing data use.

In this paper we propose a model driven optimization framework for semi-automatic road extraction from digital images. Semi-automatic means that a road is extracted automatically after some seed points have been given coarsely by the human operator, through activation of a mouse using a convenient interactive image-graphics user interface. In the model driven optimization framework, a road is represented by a generic road model that specifies both photometric and geometric constraints and defines an objective function which embodies a notion of the 'best road segment'. Then the problem of road extraction is treated as one of evaluating the objective function and generating the optimal fit of the model to the image data. Two different techniques, based on dynamic programming and least squares principles respectively, are discussed in the paper. With dynamic programming, the optimization problem is set up as a discrete multistage decision process and is solved by a 'time-delayed' algorithm. It ensures global optimality, is numerically stable and allows for hard constraints to be enforced on the solution. In the least squares approach, we combined three types of observation equations, representing the photometric part of the road model, the geometric part (modeled by B-spline) and the boundary constraints defined by operator-given seed points. The solution is obtained by solving a pair of independent normal equations to estimate the parameters of the spline. Therefore this new snake concept is called 'LSB-snakes' (least squares B-spline snakes). The issues related to the mathematical modeling and the practical implementation of both methods are discussed and experimental results of the different approaches are shown.

A system for visual road recognition in far look-ahead distance, implemented in the autonomous road vehicle VaMP (a passenger car), is described. Visual cues of a road in a video image are the bright lane markings and the edges formed at the road borders. In a distance of more than 100 m, the most relevant road cue is the homogeneous road area, limited by the two border edges. These cues can be detected by the image processing module KRONOS applying edge detection techniques and areal 2D segmentation based on resolution triangles (analogous to a resolution pyramid). An estimation process performs an update of a state vector, which describes spatial road shape and vehicle orientation relative to the road. This state vector is estimated every 40 ms by exploiting knowledge about the vehicle movement (spatio-temporal model of vehicle dynamics) and the road design rules (clothoidal segments). Kalman filter techniques are applied to obtain an optimal estimate of the state vector by evaluating the measurements of the road border positions in the image sequence taken by a set of CCD cameras. The road consists of segments with piecewise constant curvature parameters. The borders between these segments can be detected by applying methods which have been developed for detection of discontinuities during time-discrete measurements. The road recognition system has been tested in autonomous rides with VaMP on public Autobahnen in real traffic at speeds up to 130 km/h.

Within the framework of the French topographic database production, IGN² is working on problems related to the photogrammetric data capture automation. After a general presentation of the method we use for road extraction, this paper presents the use of this method within a semi-automatic approach, not very ambitious in terms of automation level but looking for a short term efficiency. A prototype has been implemented, taking into account the operational constraints, and a large part of the paper is dedicated to the evaluation of this method, evaluation in terms of geometric quality of the result and of data capture time compared with a manual data capture.

This research is part of the crucial project which aims at automating the data capture of the IGN topographic database. Our own purpose within this project is to use external knowledge provided by a G.I.S. the accuracy of which is about 20 meters to drive the road extraction in the aerial image with a one meter accuracy. The external knowledge will be integrated in a knowledge-based system. The crucial point of our work is to acquire the necessary knowledge in order to implement efficient interpretation strategies. This paper deals with the general architecture of the system and describes more especially the knowledge acquisition.

Measurements of dimensions of inaccessible objects remains a challenge in spite of new technologies available. The authors of this article approach the problem with a photogrammetric method using a laser-camera combination installed in a helicopter. The method was applied to the topographic measurement of ice fields.

This paper presents an adaptive shading algorithm based on the wavelet transform. The mathematical relations between gradient spaces (p,q) and wavelet transform subbands are investigated. A set of formulas are provided to calculate the surface gradients from the wavelets. The DEM wavelet transform pyramid provides coarse to fine terrain structures from which the aligned regional main direction (RMD) and the local detail direction (LDD) of the terrain are automatically obtained. Based on the aligned direction of the terrain, the so-called regional main light sources (RMLs) and the local additional light sources (LALs) are automatically determined. The RML gives the regional main illumination corresponding to the RMD to provide a global shading situation, while the LAL gives the local additional illumination corresponding to the LDD to provide a detail shaded relief. The final shaded image is then obtained by a linear combination of the two shaded images with respect to the RMLs and LALs. Experimental results of the shaded images from Mars DEMs show that the proposed algorithm provides optimally shaded relief images.

Methods for extracting in real time the surface structure in a certain look-ahead range from image sequences while driving are reviewed. Low spatial frequency elevation profiles of roads are the only ones successfully recovered in real-time up to now. Low frequency means that the characteristic geometrical dimension for curvature estimation is at least an order of magnitude larger than the axle distance of the vehicle carrying the camera. For the more general case of a surface with relatively high spatial frequencies an approach for real-time recognition is proposed which separates low, medium and high spatial frequency components; these components are estimated in a sequential, iteratively parallel manner. The formulation of the approach is given.

The complete object-oriented approach is generalized for the photogrammetry-oriented application development. The advantages of the object-oriented data representation and management are discussed. It is shown that the object-oriented programming is very attractive for data simulation, data unification, and fractal data representation in the digital photogrammetry software design. The original frame paradigm for object-oriented data management is described. The structured processing model is introduced to define the application area of this conception. It is proved that any full photogrammetric processing cycle satisfies this model.

The main body of remote sensing data is obtained with the aid of optoelectronic and photographic devices. This data is usually referred to as the video information since it may be presented as images of terrestrial surface on a satellite track or an airway. This is the reason of increasing interest of specialists in the field of the remote sensing devices design to the methods of synthesis of optimal data processing hardware. The design of effective systems of the remote sensing data formation and transmission are impossible without using the state-of- the-art synthesis methods of digital image processing systems, taking account of a message source and their recipient characteristic properties. It is possible to take account of these characteristic properties only on the basis of optimal N-dimensional digital filtering. From this point of view the N-dimensional filter, used for video images filtering, becomes optimal only in the case of coincidence of the pass band region of its spatial frequency response (SFR) with the isoenergetic surface of the image spectrum with allowance for eyesight characteristics. In the light of the above the problem of N-dimensional digital filters design with the given pass band region configuration becomes actual. Incidentally the practicable interest presents first of all the methods, allowing with relatively low hardware expenses to design structures, from one part operating in the real time, and from the other -- approaching best of all the given characteristics. In this case it is necessary to ensure stability during their operation. In the following we shall present the results of the synthesis method development of N-dimensional digital filters with the guaranteed stability and the given pass band region configuration, realizing the image processing in the real time.

The methods for scene reconstruction based on a series of photos of natural scene and given a priori parametric object library are described. Methods of parameters identification, means for parametric models creation, and types of features used for parameters identification are discussed.

The problem of determining spatial position and orientation of several cameras, knowing corresponding coordinates obtained by perspective projections onto the camera planes, is considered. Input data for calibration also include distances between some points in space. The calibration is carried out in two stages. In the first stage, position and orientation for pairs of images (stereo pairs) are determined. Every image is calibrated being included in one calibrated stereo pair. A special tree-like structure is built up as a result of the first stage. This structure contains input data and some information about links between the images. The calibration parameters obtained after each stereo pair calibration are considered as an initial approximation for the second stage. In this stage, the simultaneous calibration of all the images is performed to provide consistency and compatibility of final results. The proposed approach permits the user to avoid possible conflicts between calibration parameters and alleviates the problem of obtaining a good initial approximation for simultaneous calibration of multiple cameras. The experiments with real images produced promising results.

The goal is to describe a computer stereo vision system based on the well-known edge detection principles. The system provides surface recovering well enough both for far and close objects using two photos of the surface.

The Geographic National Institute (GNI) map in the scale of 1:25,000 is made out of four color plates. The green one represents the vegetation, the blue one stands for hydrography, the black is for road and toponymy, and the orange color for other various information. Our work is based on the scanned green plate. The purpose of our study is to segment and extract different zones of vegetation so as to integrate them in a GIS (geographical information system). The image is divided into untextured and textured solid areas. On this kind of image, two categories of textured regions can be found: macrotexture, represented by a set of similar graphic elements and microtexture which is represented by a set of small points or thin lines located very close to each other. There is also a superposition of micro and macro texture. These textures are synthetic, regular and often periodic. The major issue is to find a general method which can be applied to different representations. Our segmentation method is driven by the dichotomy decision. Cartographic redaction implies some visual perception rules. To be informative, variations on maps are limited: intensity, coarseness, orientation. We work on a hierarchical classification of the different areas: textured or nontextured regions, coarse or fine textures as well as intensity, orientation and periodicity of textures. Operators are associated to this hierarchy. We use statistical approach for microtextures and structural approach for macrotextures. Different texture analysis methods are tested and some have to be modified to be adapted to this kind of image. In fact, even if the image has 256 gray levels, only few gray levels are informative. For graphic elements, we focus on geometry properties. The other problem is to detect boundaries among areas. The limits are drawn sometimes but often vegetation regions are bordered by roads or rivers which are present on the other color plates. That is why, in case of the subjective boundaries, we need information from other plates to confirm our results. We noticed that for fine textures, the edge could be relieved. The first experimental tests have demonstrated that our procedure is efficient.

The paper outlines peculiarities of taking three-dimensional information from the two- dimensional image and its technical support. The task of the precise measurement of sizes and the mutual location of components of the object takes place in many fields of science and technique. Common ways of measuring by point touching use a fixed measuring probe. The fixed probe is connected with the measuring system. The free measuring probe can operate without any mechanical connection with the measuring system to transfer coordinate data about bench-mark points of the probe. The benchmark field must be situated within eyeshot of TV sensors. It is enough to compute coordinates of the probe's endpoint. This is a principle difference of the free probe measuring method from common methods. The calculation shows that the represented TV measuring system with the free probe allows the user to define quite precise coordinate information from the object to be measured.

Capabilities of the brand new method of object orientation determination are considered in this work. An object orientation is determined via determination of coordinates of special points bound to an object. Determination of points' coordinates becomes possible by virtue of an information field formed with aid of laser rays scanning in a space by a certain way. Object orientation characteristics are calculated based on measured coordinates of these points. Precision parameters of this method depend on characteristics of formed information field, quantity, and mutual placement of an object bench-mark points. High information field scanning frequency allows us to make a lot of measurements of points' coordinates that allow us to reduce errors influence. Using this method allows us to simplify and cheapen the measuring facilities essentially because it does not have to use expensive high-precision elements for an object orientation determination. Besides, the described method may have a very extensive purpose where it needs to determine orientation of an object without mechanical interaction.

Detailed investigations of the projection moire and grid projection methods used for the computer aided studies of the shape of the muscle-osseous system and postural deformities are presented. The shape determination process is readily automated using the phase methods for analyzing the fringe patterns. The issue of subtracting the reference surface inherent to the measurement-by-comparison methods is discussed. Specialized medical software for studying the deformity of the back and front of the human body illustrated by clinical examples is presented.

Deformations of a plate under different loading conditions were determined using remote sensing techniques. Imagery of the plate under investigation, acquired with a TV-camera based system, was analyzed, in order to measure electronically the average radiance over the plate through the digital numbers in each pixel. Moving from the knowledge of the radiant energy, it was possible to determine deformations, directly related to radiance, analyzing pixel-by- pixel image variations. Radiometric and geometrical considerations allowed determining deflections of the plate, after introducing a correction factor, to consider the response of the acquisition system. They were found in good accordance with those resulting from the use of traditional methods, and proved the working of the proposed technique.

Previous work by the authors showed that the endurance limit of specimens, or mechanical components, can be predicted using thermal infrared imagery. The new technique enables the determination of the fatigue strength limit in a comparatively short period of time (few thousands cycles), and using very few specimens (theoretically only 1). The present work applies this technique to rotating-bending test specimens of austempered ductile iron, an alloy whose fatigue limit is, due to the high scatter dispersion of the data points and the long testing period required, generally difficult to determine by the traditional technique. This material exhibited higher fatigue strength than the familiar nodular cast iron. This was confirmed by the results derived from the traditional Wohler test and the new technique, and supported by the data gathered from literature.