The method of gaining visual sharpness is offered in system with compression of visual data based of discrete cosine transformation (DCT). The scheme of an encoder DCT is comply with JPEG system in the following modifications: the preprocessor of chrominance performs reducing of amount chrominally different components in the ratio 4:2:2; in the block DCT the selection of transformation factors is carried out by threshold method. Obviously, that the extension of frequency bands chrominally different components allows to improve sharpness and exactitude color transmission of small-sized details of saturated colors. The use of a threshold method of factor’s selection keeps high-frequency components, which values exceed a threshold P, and the exactitude of reproduction of thin structures of an image and effectiveness of compression will depend on magnitude of a selected threshold. The functional assotiation of minimum contrast (in average color thresholds ACT) a dashed lines is obtained by methods of computer engineering, at which it is still reproduced on an exit of the decoder, from a value of a threshold P. During the modeling the transformation of blocks by a size of 8*8 elements and evaluation of minimum contrast in contrastingly equal coordinate system of Vishetsky was used.
From obtained results follows, that the sharpness of an image with magnification P is reduced and thus the distortions do not exceed admissible values at P < 4. The distortions of color transmission small sized objects in a test image do not exceed the average 3-4 ACT at working of details with a size not less (2x2) - (3x3) elements of expansion and value of a threshold P < 4. The sharpness of the boundaries accords to average value of color passages duration <2-3 and practically depends on magnitude of a selected threshold a little. The results of an experimental researches have shown, that on the average expenditures on one element do not exceed 2-3 bits/component (at P < 4) and thus good quality of an image with subjective evaluations "well" and "perfectly" is ensured.
A successful solving of the remote eco-monitoring tasks is based on the timely and detailed analysis of the multi-spatial data of a different content. A great importance is given to elaboration of the non-traditional approach, which provides the following:
1. A comparison of the ecological data about condition of the objects in the
2. Obtaining the full idea about the process of monitoring and its different phases.
With this purpose a cognitive graphical model of the ecomonitoring (CGME) in the form of a totality of the correlated models has been worked out. Presentation of the semantic information of these models can take the following forms: the textual form, the audio form, the graphical statistical and the dynamical images. The components and the logical structure of the model have been marked out .
The methodical bases in building the regulated m-parametric patterns of the multi-spatial objects (processes, phenomenons) have been worked out. This favours a creation of the ecological data banks on the united base. This also promotes a further harmonization of the international information systems. In order to increase the efficiency of the natural objects' observation, the method and algorithm of the computer graphical building of m-parametric patterns of the airborne spectrometers' characteristics was worked out for the first time.
The pattern developed on the base of the local signs serves as a cognitive mean of the simultaneous reinforcement of the collective informativity of these signs. CGME allows to cover quickly relations between the more wide range of the objects and phenomenons, to choose the more considerable of them and to recognize the logical line of the problem. All of the mentioned above favours decreasing of the risk factor during the ecoexpertise.
The cognitive computer graphics is used in science for visualization fundamental abstractions with the aim of activation of creative (intuitive) thinking of scientists. The new information and multimedia technology gives many new instruments and based on this instruments methods for cognitive visualization.
This work is associated with research of new multimedia instruments abilities for cognitive computer graphics tasks. The problem area is relation algebra. There are two types of objects in graphics for operation - color and form. That is why we done two cognitive models - color and form. In color model the form is fixed and vice versa. The models were done on Microsoft PowerPoint in form of slides sequence. Every slide shows the one of fundamental parts of relation algebra (attribute, cortege, operations, etc.). Animation and sounds are used in many slides.
The research shows that the Microsoft PowerPoint is a simple instrument and its abilities are not effective ones for cognitive computer graphics aims. The animation instruments, such as AmiPro, 3D Studio Max etc., are more effective for cognitive computer graphics models creations. But in this case the creation of model is more complex. We think that the known multimedia and graphic instruments couldn’t used for cognitive computer graphics models creations and deal with it. Special instrument for this purpose will be creative. The variant of this instrument is discussed in this work.
Graphical Information Support Library (GISL) is intended for use of some opportunities on organization of representation the diagrams in other appendices. In particular, the following opportunities are included in it:
- Organization of a graphic window in a window of the user;
- Creation of simply graphic elements from the following list: a rectangular, rhombus, ellipse, polygon, arrow (line), broken arrow (line);
- Support of the user types given for graphic elements;
- Creation multilayer and multiscreen images;
- Scaling the image;
- Preservation of the image in a format BMP.
The library is written in language Borland C ++ 5.02, consists of three heading files with expansions .H and file of static library with expansion .LIB and is used while only by connection of a file .LIB to the project Borland C ++ 5.02. The library can be used for creation of the simply graphic editors, reception of figures both from the program, and from the screen, organization of graphic information structures like hypertext with use of several screens of the images, etc.
At an exhibition the demonstration program showing some opportunities of library is submitted. Besides, the library is applied in the shaper of question-answer structures "WIQA for Windows 95" (PERT-graphic, Petri net), also submitted on an exhibition.
The expansion of opportunities of library (increase of quantity graphic primitives, introduction of the BMP-images, etc) and representation it as DLL-library and components for use in Borland C ++ Builder.