It is very difficult to identify the birds by human and computer because different variation of birds to control and analysis. We need to identify Birds that weigh over 1.8 pounds (0.816 kg) because these birds strike can cause lots of damage to the aircraft. In this research we have used Caltech-UCSD Birds-200- 2011 Dataset. It is an extended version of the CUB-200 dataset. In this dataset there are 200 bird species category, In this research project we create handled device (android application). In this handled device we use transfer learning to train this model. We try to different model to get accuracy such as VGG16,VGG19, MobileNet,ResNet50,InceptionV3 and Exception.