With these foundations, a set of anomaly recognition methods is proposed, which is supported by statistics, fractal geometry, artificial neural network and computer technique.

To extract the thermal infrared(TIR) anomaly information related to the tectonic activities from remote sensing data,a new pre-earthquake TIR anomaly recognition method and quantitative analysis model based on reference field are presented.

The study of metallogenic anomaly identification based on deep soil data with multifractal method——A case study of Cu element in Tongling ore clustering area

In this paper,the author shows solving process of two characteristic lines and analyzes its corresponding significance of geology and geophysics and describes results for anomaly identification through a model and a field examples.

Hierarchical multi-observation model for multi-person activity recognition

Human action understanding is generally divided into three fundamental sub-processes:feature extraction and motion representation,activity recognition and higher level activity and scene understanding.

An activity recognition method of multiple view angles from monocular videos was introduced.

Spam behavior recognition based on data mining in session layer;

Research on Key Technology of Spam Behavior Recognition Based on Data Mining Technology;

Behavior Recognition Method Based on Multi-camera Information Integration