An Anomaly Analysis Method Based on Morphological Features

The basic method and concept of time series feature extraction is applied to anomaly analysis of seismic data. A new method for piecewise representation based on morphological feature points and a new feature extraction method are proposed. The experimental results show that the proposed piecewise representation method can achieve smaller fitting error and retain the morphological features of the original data better than existing approaches when applied to three real world datasets. The experimental results also illustrate that the proposed method for feature extraction can extract changes in electromagnetic data and distinguish the extent of abnormal changes in real world datasets. In one case, the results show that the approach could identify change features that might be related to an earthquake.