|
Building And Hosting A Computer Vision API On AWS Using An EC2 Instance
[Full Text]
AUTHOR(S)
Kaushik Arvind Jadhav
KEYWORDS
API, AWS, Cloud Computing, Computer Vision, EC2, Flask, Object Detection
ABSTRACT
This paper aims to propose a methodology to host a Computer Vision API that can recognize and classify various objects in a particular image on Amazon Web Services (AWS) using Amazon Elastic Cloud Compute (EC2) instances. The images that were used to train our deep learning model were collected from different sources. The API takes an input image from the user and generates an output of the object present in the image and its most abundant color. In order to run Python scripts via the web, the RESTful API protocol has been used. Before developing the API, a Web App was developed and tested locally using Flask. The final end product is able to detect objects present in images and determine the most abundant color in the frame of the image.
REFERENCES
[1] Paul Richard, Alexandru Telea, Alain Tremeau: Computer Vision, Imaging and Computer Graphics Theory and Applications. Springer, Portugal (2018)
[2] Suhas Athani and C.H.Tejeshwar: Face Identification and Face Classification Using Computer Vision and Machine Learning Algorithm. Springer, India (2017)
[3] Debalina Barik, Manik Mondal: Object identification for computer vision using image segmentation. IEEE, India (2010)
[4] Kanza Azhar, Fiza Murtaza, Muhammad Haroon Yousaf, Hafiz Adnan Habib: Computer vision based detection and localization of potholes in asphalt pavement images. IEEE, Canada (2016)
[5] Chin-Yun Hsieh, Hong-An Hsieh, Yu Chin Cheng: A method for web application data migration based on RESTful API: A case study of ezScrum. IEEE, Japan (2016)
[6] Erik AlbertSudarshan, S. Chawathe: Deploying a Multi-interface RESTful Application in the Cloud. Springer, USA (2013) J. Williams, “Narrow-Band Analyzer,” PhD dissertation, Dept. of Electrical Eng., Harvard Univ., Cambridge, Mass., 1993. (Thesis or dissertation)
|
