In work described generalized algorithms of an immune network, 
and also the architecture of an immune network developed by authors 
and functional blocks included in it. The structure of classes 
of the tool environment for development of artificial immune networks 
of the problems of classification intended for the decision and recognition 
of images is shown. By the example of test problems the investigated role 
and value of some parameters of immune system on quality of the decision 
of problems of classification.