Font Size: a A A

Image Recognition By Multiple Classifer

Posted on:2018-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:M W WangFull Text:PDF
GTID:2348330518996538Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
In recent years,with the development of sensor imagery technology,we can obtain lots of sensor imagery information. But how to handle the information is also important. The traditional methods for classification is not only inaccurate but also producing lots of redundancy.This article mainly research the problem of sensor imagery classification,and uses the band information of the sensor imagery as the features. The classification method is based on multiple classifer and uses three layers of rule to make the system reliable. This article gets 20 sensor imagery from different Latitude and longitude for recognition and compare the new methods with the traditional methods. At the end, this article combines the classifer with the wireless transmission system to decrease the communication traffic.
Keywords/Search Tags:satellite sensor imagery, spectrum features, classifier combination, the neural network, SVM, deep-first search
PDF Full Text Request
Related items