BIRDS BREED CLASSIFICATION SYSTEM USING CNN

Now a day some bird species are being found rarely and if classification of prediction is difficult. Naturally, birds present in various scenarios appear different sizes, shapes, colors, angles from human perspective. Besides, the images strong variations to identify more than audio classification. Also, ability recognize through understandable. So, this method uses Caltech-UCSD Birds 200 [CUB-200-2011] dataset for training as well testing purpose. By using deep convolutional neural network (DCNN) algorithm an image converted into grey scale format generate autograph by tensor flow, where multiple nodes comparison generated. These compared with score sheet obtained it. After analyzing sheet, it can predicate required highest score. Experimental analysis on (i.e., [CUB-200-2011]) shows that achieves anaccuracy identification between 80% 90%. The experimental study done Ubuntu 16.04 operating system Tensor flow library. KEYWORDS: Caltech-UCSD; pixels; TensorFlow.