January 16, 2019 Dave Desmarais

Photo Based Clothing Measurements

The garment industry adheres to a specific set of rules for measuring clothing; they include how to take the measurements themselves and the error-tolerances allowed during design and manufacturing. Measuring clothing is a process that, until now, has been reserved for humans. Through machine learning and computer vision, Tailored has introduced a system of capturing garment measurements via photo. Photo based clothing measurements, at present, are as accurate as a trained human and will only get better. This coupled with drastic increases in speed and consistency and the ability to capture more data are a few of many reasons clothing retailers are seeking out this technology.

Speed

There are three major apparel-retail sectors and they each have processes where the time spent during certain tasks can be drastically reduced with the integration of automated garment measuring.

1. Traditional Retail

In traditional retail, due to the physical limitations of human beings, only 2-5% of a production run is physically measured against design spec during quality inspection. With photo based clothing measurements, the time required to capture the critical measurements of a garment is reduced from an average of three minutes to 20 seconds or less. For a retailer, this means an increase in the number of garments spot-checked in the same amount of time, or a drastic decrease in the resources put towards this task if sticking with the same 2-5%.

There is also the issue of physically transferring a sample garment from the factory to the retailer’s HQ to inspect the measurements. This is done because the factory will always measure in a way that benefits them, meaning measuring as close to design spec as they can, even if the actual measurements are outside of tolerance. Replace this process with the ability to snap a photo of the garment while it’s at the factory, and instantly see a close-up image as well as all the Points of Measure, and you have just reduced your time from weeks to minutes.

Photo based clothing measurements

2. Re-commerce

In the re-commerce sector, the major players have massive distribution centers scattered across the globe. These dc’s accept tens of thousands of garments daily which must be inspected and hand measured before uploading to the web for sale. Photo based clothing measurements not only speeds up this process, but sets the retailer up for a simple segue into a consumer solution using the same photo-based technology. For example, imagine you’re shopping on a second hand clothing site and you’re not sure which of the thousands of garments might fit you best? Just snap a picture of a well-fitting garment you already own, and the site will instantly recommend garments with the same size and fit as your garment at home.

3. Monthly Subscription/Box Services

Monthly subscription retailers see a similar onboarding issue as re-commerce companies. Most monthly-box retailers do not produce the apparel they send to their customers, and if they do, it’s less than 40% of their overall available product. The ability of a stylist to successfully pair their customers with properly sized garb means they must have an accurate representation of the physical dimensions of the garments they offer. This means hand-measuring the apparel shipped to them by the producing retailers, and also means the same human limitations.

One Standard of Measure

Consistency in garment measuring is major gripe had by all apparel retailers. Fashion is a global business and we all understand that most, if not all, of the apparel sold in a particular country is not actually produced there. This along with the absence of some form of standardized measuring leads to high levels of inconsistency in the business. Using software to measure clothing is like measuring all garments with the same hands, no matter where in the world the measuring is taking place.

Besides the obvious internal use cases, photo based clothing measurements is the next evolution in e-commerce. Consumer confidence is a big push as retailers think of new ways to drive shoppers to their sites and how to help them choose the right size/style the first time. Sure virtual fitting rooms are a nice gimmick, as is body scanning, but nothing will tell a retailer more about how a person likes their clothing to fit than the clothing they already wear. People are generally comfortable with that they know, and snapping a photo of a well-fitting garment one already owns to see recommendations when shopping for new clothing allows for a confidence level we have not seen before in online shopping.