Abstract
Several works have been done in the area of image retrieval systems, and many are still trying to provide improvements for a better model for retrieving said images. Image segmentation using clustering techniques is one of the most used approaches. There are various clustering methods available, but the non-linear k-means clustering technique is the most used method. In the following research, a model of retrieving images using a non-linear classifier aided with a convolutional neural network is proposed. Both algorithms were exploited and paired in terms of feature extraction and classification. Comprehensive evaluations over a dataset containing over 7,000 pill images of 1,000 pill types obtained from the National Library of Medicine database demonstrate significant success during the data classification using the proposed model.
Publication Date
12-2020
Document Type
Master's Project
Student Type
Graduate
Degree Name
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research (Dubai)
Advisor
Sanjay Modak
Advisor/Committee Member
Khalil Al Hussaeni
Recommended Citation
Adewumi, Ezekiel Adebayo, "A Convolutional Neural Network (CNN) based Pill Image Retrieval System" (2020). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/10730
Campus
RIT Dubai
