How to Build a Monitoring Application With the Google Cloud Vision API

The nice part about the samples that Google has put up for the Cloud Vision API is the range of applications and platforms. They range from processing files in Google Cloud Storage, to Text Detection and also how this API could be employed in your Android and iOS applications today. oth mobile samples allow you to pick an image from your phone and analyze the image for feature detection. You could also easily extend it to sort out all your gallery photos, take a picture while traveling and retrieve its landmark information (if there is one) and so on. The possibilities are limitless.

Interesting Use Cases

Since its Alpha and now Beta program, the Google Cloud Vision has seen developers not just explore its capabilities but  also build out Proof of Concept applications that use it in interesting ways.

One such use case is from Google Developer Expert, Ivan Kutil, who analysed city traffic using Google Cloud Vision API. In this application, Ivan processed the image grabs from a street and subjected the images to LABEL_DETECTION. He mainly looked for features like “road”, “traffic” and the relation between them. 

google cloud vision api city traffic application


Google Cloud Vision API is currently in Beta and the team has come up with pricing per feature request that you ask for in an image. Earlier in the article, we looked at various Feature Requests that the API supports like LANDMARK_DETECTION, LOGO_DETECTION and others. Each of the Feature Requests that you invoke in an API for a given image is counted as a single unit. For example, on a single image, if you invoke FACE_DETECTION and LANDMARK_DETECTION, then it will be counted as 2 units.

At this point in time, there is a free tier, where you are allowed 1,000 units per Feature Request per month free. Beyond that there is a tiered pricing model based on the number of units that you use in a month.

Check out the pricing page for more details. It is interesting to note that LABEL_DETECTION is priced higher than all the other Feature Requests.

Since the API is currently in Beta, it is subject to breaking changes and is not governed by a SLA or deprecation policy. So it is generally advised to not consider it for production yet.

Machine Vision API Vendors

Google Cloud Vision API is not the only API available to employ machine vision into your applications today. Both IBM and Microsoft offer APIs in this space and it would be worthwhile to check them out too.

IBM, as part of its acquisition of AlchemyAPI, offers Face Detection API as part of its overall Watson Suite. Similarly, Microsoft has Project Oxford, which offers Computer Vision API, Face API, Emotion API and more.


Artificial Intelligence is rapidly making its way into mainstream computing. The ability to integrate powerful Machine Learning Models into applications, gives them capabilities that were not possible before or were assumed to be many years into the future. With the release of Machine Vision APIs like Google Cloud Vision, it opens up a whole new stream of applications and while it is early days here, it is a foregone conclusion that as more usage and data is generated, these models will become more powerful and would bring up increasing accuracy to your applications. 

Be sure to read the next Machine Learning article: An Analysis of Brexit With the MonkeyLearn Machine Learning API