[Pattern recognition]. (Greece) ritis <version 4> Note 7 _ feature generation (2)

This chapter focuses on feature generation in image analysis.

1. Regional Features: ① texture properties ② local linear transformation extracting texture features ③ moment ④ Parameter Model

2. shape and size Features: ① Fourier features ② chain code, which is the most common technique for describing boundary shapes ③ moment-based (moment) features ④ geometric features

3, fractals (a glimpse at fractals)

Two important tools were introduced into the pattern recognition application field in the 1980 s:Neural Networks and Wavelet Analysis

In the same period,Fragment and fragment dimensionAlso introduced.

4. Typical features of speech and audio classification

[Pattern recognition]. (Greece) ritis <version 4> Note 7 _ feature generation (2)