Text Extraction from Images Using Cloud Services


Aditi Narkhede, Manswi Meghe, Ujwal Rahile, Vaibhav Dhawale, Rakhi Wajgi
Department of Computer Technology, Yeshwantrao Chavan College of Engineering Nagpur, India.
Ashish Bhagat, Research Consultant
Jawaharlal Nehru Medical College, DMIMS, Sawangi (M), Wardha, Belgaum, Karnataka, India.


Nowadays everyone has mobile phones and people usually capture images of textual documents, handwritten papers,banners, packed food, prescriptions,bills and anything which is of their work. Sometimes they need to work on same data which is present in images and they write/type it manually which requires a lot of time. Also, there is huge demand of storing the data or information available in paper documents format into a digital format so that the data is searchable and editable. Hence there is need of software which is able to extract text from captured images. So we are developing an android application which can extract text from uploaded images. The uploaded image is analyzed using AWS rekognition service of amazon web services which is implemented using various machine learning algorithm which convert the detected text into machine readable text.This AWS recognition is a cloud-based service which allow developers to directly use the machine learning algorithms without actually knowing about it which is a feature of platform as a service (PaaS) of Cloud Computing . With this service we can built an application using machine learning concepts without prior knowledge of it, but by simply accessing this service through cloud.The main advantage of this application is that it is free of cost as we are using all the services by remaining in a free-tier only. The objective of this paper is to compare different methods used for extracting text from images with the method we are using and describing our method of implementation for recognizing the text from image provided for better understanding of the reader by using particular sequence of different processes.