Comparison of two approaches to the construction of sets of linear local features of digital signals

The paper compares two approaches to the construction of sets of linear local features (LLF) of digital signals. The first of the analyzed approaches forms a set of LLF using independent efficient LLFs, each of which has its own algorithm for computing the feature value. The second approach constructs an efficient set of LLF, which has a single algorithm of joint computation of values of all features. Analytical and experimental comparison is made for several indicators of the computational and qualitative properties of the constructed LLF. An experimental comparison of two approaches with known solutions is also presented.