Exercise using Software as a Service: Google Cloud Vision API

due start of class Oct. 24

GROUP work

Your groups your will be developing code AND documentation for a client HTML form (you must also create) that uploads an image from browser client and invoke a servier side program that recieves the image data and sends it to the Google Cloud Vision Labeling service and takes the resulting label annotations and returns an HTML response to the browser client with a welll formatted list of detected labels. SERVER SIDE code (NodeJS+Express) must be deployed to Google App Engine. Create a shared GitHub repository to host your code and a wiki documenting the process and code. Turn in URL of the github repository to Canvas->Assignments-> Exercise SaaS

TIP1: To call Google Cloud Vision api: I would probably do with the client libraries rather than the REST api https://cloud.google.com/vision/docs/libraries

• Google documentation (search if link broken) on how to download Google Cloud Vision Client Library

 

TIP 2: search on image file upload and creating form for user on client browser to specify image file to be sent to server. Look online for sample code for NodeJS+Express to recieve image files (one example of many)

<form action="server_program_toUploadImageFile" encType="multipart/form-data"> <input type="file" name="pic" accept="image/*"> <input type="submit"> </form>

 

Deliverables

• Create a Github for your team that has the NodeJS+Express code and wiki that documents the work you have done with the following wiki pages:

  • app - link to the deployed application (that is up and running for me to try)

  • demonstration - screen shots showing app working (before and after results)

  • explanation - discuss in detail the code and how it works and how you deployed it to GAE

  • state of system - tell me if fully working or if any issues.

• Turn in URL to github (once per group) on Canvas->Assignments>Exercise-SaaS)