The problem Colorage solves

We all love to shop wardrobes online. But on any e-commerce website, when we apply filters to customize the results according to our taste, the outcome is not always satisfactory, especially when sorting according to color. Retailers selling their products on these sites often don't provide accurate and relevant tags. Let's say an apparel containing several colors in its description (for example: white and blue), but having a picture clearly showcasing a blue shirt. While searching for a white shirt this apparel may show up and may even rank very high, because the relevant color – white, is present in the description. However correct this may be, having a number of images showing a blue object while searching for a white one does not give the best impression of relevance to the end-user.

Many apps and scripts exist already which can extract the dominant color in the image, but they will fail miserably in our use case because the dominant color in the images used for showcasing clothing often has the background as the dominant color. Too much skin in the image will also cause detection problems.

Our idea is to develop a script that can easily be integrated by the e-commerce websites, which can automatically tag the apparel color, but without the use of any deep learning frameworks, because of their requirement of large datasets and heavy computation resources, by using a combination of simple image processing and computer vision techniques. It will provide a quick, easy way for users to enrich their records a bit and improve their ranking, especially if they have to deal with poor descriptions in their records.

Challenges we ran into

Challenges in running the API, Handling full size images, skin removal, Background and forground seperation