National Flag Recognition Using CNN

Abstract

This thesis is on “NATIONAL FLAG RECOGNITION OF USING CNN”. Present work suggests a procedure for recognition of national flags for the countries of SAARC for their digital image. Arithmetical structures take out from the color stations that are joint together to create the feature vector for discernment among the national flag. For low accuracy, time consuming, and very slow performance we use CNN classification to recognized national flag of all countries of world. Albeit advances and vivid application of Artificial Intelligence and Computer Vision in different national flags of all countries of whole world there contains low effort in the application of computer vision recognition. With a goal to successfully applying computer vision techniques to predict a country name based on national flag image. This paper describes a novel Deep leaning based approach for recognizing national flag. we have used inception v3 and achieved an average of 96.6% accuracy rate which is the best of all previous.in Our data set, here we use 4000 images to test the performance of our classifier.