Conditions of the program execution

Automatic vectorization is performed on the basis of analysing the distribution of brightness of image pixels.  For determining the statistical characteristics of the brightness distribution the user digitizes patterns - the regions in the picture belonging to a recognizable object. These data are used in the further at classification - definition of a belonging of pixels to recognizable object.

This approach is based on the assumption that recognizable objects are displayed in the photo by the same color as in the template. In general, this assumption is incorrect. For example, the image of road can be closed by the objects which color is distinct from color of the road - crones of trees, shadows, a road marking, cars. If there is a lot of such regions in a picture, the result of recognition will be unsatisfactory, a considerable amount of time will be required for editing results of an automatic vectorization. Therefore the automatic vectorization gives increase of productivity compared to manual vectorization in regions where the above-mentioned hindrances are minimized.

For acceleration of processing, in operative memory the files for storage of intermediate data are created. It imposes restriction on the size of processed rasters. It is not recommended to process rasters larger than 300 Mb.

The complex of the auto decryption and vectoring is the 64 bit application, therefore it can be carried out only in 64-bit operating system (since Windows 7 х64) on computers with architecture of Intel processors (Core i5, Core i7 or is more senior).

Recommended requirements to the hardware and software:

- RAM: 4 GB or higher;

- the volume of the hard drive: 1 TB or more;

- screen size of at least 1024 by 768 pixels.

The program is protected by a dongle. Processing of rasters smaller than 2500х2500 pixels in a demonstration mode (without the dongle) is allowed).