AdaBoost based characteristic extracting method for pattern recognition

Abstract

The present invention includes 1, determining two kinds of sample, according to inputting sample data, extracting high dimensional selected features; 2, determining one weak sorter assembly relative to single features training, each weak sorter in said assembly corresponding one feature; 3, respectively setting initial weight(ed) value for said two kinds training sample configuring; 4, according to current weight(ed) value inputting training sample to proceed training; 5, according to training result, selecting minimal error ratio preset number weak sorter corresponded feature as current turn feature selection result; 6, updating all training sample weight(ed) value, re-executing 4-6 steps until ending preset turn. The present invention solves high store content and calculated amount problem in current selected method.

Description

technical field [0001] The invention relates to a feature extraction method, in particular to an AdaBoost-based feature extraction method applied to pattern recognition. Background technique [0002] In the pattern recognition problem, the basic task of feature selection and extraction is how to find out the most effective features from many features, analyze the effectiveness of various features and select the most representative features. In this way, the complexity of the classifier can be reduced, the storage space of the features can be reduced, and features with poor classification ability or even no classification ability can be eliminated to improve the accuracy of the classification algorithm. [0003] Especially in the field of face recognition, feature extraction is a very critical link in face recognition. Commonly used features are grayscale features, edge features, wavelet features, Gabor features, etc. Among them, Gabor provides multi-scale and multi-directi...

AI Technical Summary

Problems solved by technology

The main problem of this type of method is that when the sample dimension is high and the number of samples is large, the calculation amount of the selection algorithm is very large, and the storage space requirement is very high, and the algorithm is not designed for the subsequent classifier, and the performance of the classifier does not directly benefit from it. The result of feature selection, which also affects the classification performance to a certain extent

Images