In the aspect of artificial intelligence, it mainly focuses on the research of artificial neural network, multi-agent system (MAS), genetic algorithm and expert system, and tries to apply these intelligent technologies to the design and development of intelligent control system.
In the aspect of pattern recognition, the main research directions are:
data mining;
clustering analysis of medical images;
target detection and tracking of video images;
semantic analysis of video events;
face recognition;
application of wavelet theory in image signal processing.
At present, with the support of NSFC and several provincial and ministerial fund projects, these intelligent technologies and pattern recognition technologies are deeply studied and applied to intelligent transportation system, content-based video / image retrieval system, virtual reality environment and other issues.
In particular, in recent years, we have cooperated with Zhenjiang Jiangbin hospital to collect CT images and Mir images in the past two years, and constructed a distributed medical image database with a total of more than 200000 images composed of nine parts, including head and limbs, according to the classification of human body parts. Based on the database, the research on content-based retrieval technology, region of interest feature extraction and data compression of medical image is carried out, and a database management system of abdominal medical image based on content is developed. At the same time, the software system of fatigue driving detection, electronic medical record writer and embedded electronic medical record system are developed. These systems can be applied in practice, and are expected to produce greater economic and social benefits.
