Efficient Gold Tree Child Items Classification System Using Deep Learning

Dr.R.S. Sabeenian, M.E. Paramasivam, R. Anand and S. Hariharan

The paradigm of world has shifted to Machine Learning. The growing trend of machine learning has brought various innovations in the field of human’s life.. Sorting of image using particle classification can’t remain aloof from the influence of machine learning. An image sorting system using Convolution Neural Network (CNN), made a breakthrough in the field of deep learning in turns of accuracy and outperformed the previous methods such as nearest neighbour, linear classifier, and SVM classifier. The proposed system sorts the children of a tree and matches them with image stored in database. The first stage is image acquisition and applying manual binning. After binning, training and testing of images are done. Database created is taking into account which has more than 300 images of different classes. Sorting is done, when the trained image tantamount with image in the database, as a final stage of proposed model. The Proposed framework utilizes AlexNet CNN technique for profound investigation which gives 97% exactness better than the current framework with SVM classifier.

Volume 12 | 04-Special Issue

Pages: 1845-1859

DOI: 10.5373/JARDCS/V12SP4/20201671