Research On Algorithms Of Remote Sensing Image Classification Based On Artificial Immune System

With the development of remote sensing technology, more and more subjects and techniques are incorporated in the field of remote sensing. Recent years, Artificial Immune System (AIS) which is inspired by biological immune system become active in many fields. For biological immune system has ability of recognition and classification, AIS for remote sensing image classification could be abstracted from immune system. Based on the analysis and generalization of status quo of international and domestic research and applications on AIS, several researches on classification of remote sensing image based on AIS are made. The major work implemented and the goals achieved are listed as follow:1. Status quo of AIS and remote sensing image classification is summarized.2. The model of Artificial Immune Recognition System (AIRS) for remote sensing image classification is built; the flow of the algorithm is illustrated; classification of image is carried out and the key parameters are analysed. For AIRS is sensitive to noise, improved AIRS is proposed to make up this deficiency.3. The method of remote sensing image classification based on Evolutionary Artificial Immune Network (EAIN) is proposed. The model of EAIN for remote sensing image classification is built and validated. The experiment results show that the algorithm is sensitive to the value of network suppression parameter.4. The model of remote sensing image classification based on Artificial Immune Network for Diversity is built using the methods of special network connections and coding; the validity of the model is tested on the benchmark dataset and real remote sensing image. The experiment results show that the precision of algorithm is affected by the vigilance paramerter.At last, the main work of this thesis is summarized and more research work that needs to be done is also brought forward.