Abstract
	At the beginning tasks of computer picture recognition were concentrated on figure recognition and signatures and only afterwards more complicated identifying tasks were solved. At present, the image analysis enables to solve even some problems connected to measurement and conducting of technological procedures. The article shows possibilities of the application of the above mentioned method for some tasks connected with material forming processes. We have in mind a computer analysis of wedge test and temperature determination of highly heated materials.In the process of determination of surface temperatures the computer colour analysis is used and artificial neural networks algorithms are used for the evaluation. Advantages of the presented method consist in good mobility and possibility of monitoring of temperature fields in places with difficult access to the operator. From measurement analyses it was possible to define convenience of various devices for the given purpose, work regions and lawful acts affecting measurement. The modification of the method mentioned above allows with sufficient accuracy reading off printed thermographs and other charts of scalar item expressed by quasi colours (e.g. tension distribution in a material, etc.). By exploitation of computer analysis of the outline of the rolled out material, the wedge test becomes very effective because assigning between corresponding cross sections of the wedge and rolled out sample is very accurate, which allows to improve considerably even calculations of deformation parameters.