A.S. Vedenina, L.M. Smirnova

Efficiency of the use the method computer plantography in medicine depends of not only accuracy of its result, but also examination labour content. Calculation of the diagnostics criterion of the foot condition is produced automatically in hardware-software plantographic complex. But the stage of the placement by operator on pantographic scene identification point precedes by this. For this operator must visual to define the borders of the foot sidebar and its supporting imprint, which are distinguished by them on pantographic scene on color tone, saturation of the colour and other, which are typical for anemias zone of skin cover of plantar surfaces of the foot under influence of the load. Efficiency computer plantography could be increased, at least once if placements process of base identification point and line is managed to relieve for operator, including creating sidebar of the supporting imprint, and more so if this process is managed completely to automate. The use possibility of the digital processing methods of pantographic scenes are studied for automation of the foot condition diagnostics under mass examination. The scenes got on module "PlantoScan" in composition hardware-software complexes "Skan" and "DiaSled-Skan" were used. Such choice was motivated that scenes digitization of plantar surfaces of the foot in these model of the equipment is founded on method of the optical scanning, which provide greater accuracy of the plantogramm reception. For separation of the scene element areas of the foot methods threshold-segmentings of the scene and clusterizations on colour are aplyed. For finding of sidebar on plantogramm scene were aplyed methods of high-frequency filtering (HFF), gradient methods (the Robert-s, Sobel-s and Previtt-s operators) and methods with use of second derived (the Laplas-s operator). On the grounds of called on work possible to judge about that positive effect is reached: variation of pair "brightness - contrast" - for increasing of plantogramm clearness ; using of HFF or Previtt-s, Sobel-s or Laplas-s filters - for accenting of the scene borders of plantar surfaces of the foot; the threshold-segmentings and clusterizations on colour - for separation and calculation of supporting imprint area. The best result may be reach by using of combination of the different methods of segmentings.