Realtime People Counter with Google’s Cloud Vision API and RxJava

Welcome to the ninth blog of 52 Technologies in 2016 blog series. Recently, Google released Cloud Vision API that enables developers to incorporate image recognition in their applications. Image Recognition allow developers to build applications that can understand content of images. Google’s Cloud Vision API is very powerful and support following features:

  1. Image categorization: The API can help classify images into categories. You can build powerful applications like Google Photos that do automatic categorization.
  2. Inappropriate content detection: The API can detect inappropriate content in an image like nudity, violence, etc. It uses Google Safe search capabilities underneath.
  3. Emotion detection: This allows you to detect happy, sad or moderate emotions in an image.
  4. Retrieve text from the image: This allows you to extract text in multiple languages from the images.
  5. Logo detection: It can help you identify product logos within an image.

There are many possible applications that you can build using this powerful API. In this tutorial, we will learn how to build a realtime people counter. The application will subscribe to a twitter stream for a topic and would return number of people found in each image. We can then use this data to get advanced statistic like number of people in a time frame using RxJava buffer capabilities.

You can read full blog https://github.com/shekhargulati/52-technologies-in-2016/blob/master/09-cloudvision/README.md

Advertisements