A comparative analysis of the computational complexity and probability of false
recognition for object identification methods in images, which are based on a
comparison with a pattern, viz., correlation-extreme method, contour analysis
method, and stochastic gradient identification algorithm. Possibilities of
parametric optimization of the studied methods are considered.