How to Use Computer Vision to Automatically Filter Images

By Michael Auderer | March 29, 2019

While image recognition is a seemingly simple task for humans to do, it’s actually a very complicated process. Computers can already “see” images in the sense that they store the data of every tiny pixel to make up the full picture. However, recognizing patterns in this data and creating a quickly searchable memory of common patterns is extremely difficult.

Fortunately, Google Cloud provides a very simple API to detect the content of images using computer vision. This can be useful to categorize images, recognize text, or even moderate content to prevent inappropriate images from being posted on your site. In this article, we will quickly build CatBook — an image sharing site which only allows pictures with cats in them, no dogs allowed!