Image Recognition of Wild Birds in Yunnan Based on Improved Mobile Deep Learning

The bird resources in Yunnan Province have great ornamental value and economic value. These variable and colorful features bring many challenges to their intelligent recognition and efficient management. The primary requirement of bird image intelligent recognition is convenience, portability and rapidity. This paper proposes MobileNet to recognize the bird images because MobileNet has smaller scale, fewer parameters and faster speed. The experiment results show MobileNet also have higher accuracy and can be applied in the bird watching and wild bird recognition. Therefore, this method has good practicability and can be generalized to the identification of other wild animals.