Bird Species Classification using multi-scale Convoluted Neural Network with Data Augmentation Techniques

  Pankaj Prakash Patil,  Atharva Dhananjay Kulkarni,  Aakash Ajay Dhembare,  Krishna Adar,  Rahul Sonkamble

Bird predation is a major problem in aquaculture. Nowadays bird Species are becoming rare, so we need to recognize them. Image recognition software can improve their efficiency in chasing birds. We proposed the System for Bird species Classification is a challenging problem due to the variation and different viewpoints of the camera. In the existing system, there are some disadvantages. We tried to overcome it by integrating the new feature into the multi-scale Convoluted Neural Network with Image Segmentation for Indian bird species classification, an algorithm is proposed to get the final classification result. Three recognition techniques were tested to identify birds i.e., image morphology, artificial neural networks, and template matching have been tested. We proposed a new feature that can improve the correct classification rate of the model as well as the accuracy of the model in the prediction of Birds classification. In this challenge, the bird image classification task, especially for Indian birds, is based on a limited but diverse set of crowd-sourced data. Especially, the present challenge involves a low amount of labelled data to build good classification approaches for effective classification. Up to now a lot of research has been done to identify bird species. Finally, we have proposed a methodology to improve accuracy in the identification of bird species.

Keywords- Data Augmentation, Dropout, TensorFlow, Keras, DT, CNN, Multiscale

Unique Identification Number - IJEDR2102047
Page Number(s) - 289-291
Pubished in - Volume 9 | Issue 2 | June 2021
DOI (Digital Object Identifier) -    http://doi.one/10.1729/Journal.27387
Publisher - IJEDR (ISSN - 2321-9939)

  Pankaj Prakash Patil,  Atharva Dhananjay Kulkarni,  Aakash Ajay Dhembare,  Krishna Adar,  Rahul Sonkamble,   "Bird Species Classification using multi-scale Convoluted Neural Network with Data Augmentation Techniques", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.9, Issue 2, pp.289-291, June 2021, Available at :http://www.ijedr.org/papers/IJEDR2102047.pdf