CASE STUDY
AN ML-BASED CARD COLLECTION AND TRADING PLATFORM TO HELP NEXTGEM BECOME THE ENTHUSIAST’S CHOICE!
ABOUT
NEXTGEM
Founded by an avid collector with an extensive collection of rare sports cards, NextGem aimed to become the platform of choice for card collectors and traders worldwide. For the founder, It was a great way to bond with his brother over their shared nostalgia as they set about acquiring all of the cards beyond their reach as teenagers.
THE PROBLEM
The company aimed to enable collectors to create digital copies of physical cards and approached us for the development of an intelligent scanning system. Without a sophisticated algorithm at work, the majority of digitized images of trading cards result in one or more of the following unwanted image artifacts or noise:
SHAKE DURING IMAGE CAPTURE
IMAGE TILT ISSUES
OBLONG OR MISSHAPEN ANGLES
MAINTAINING HIGH-QUALITY IMAGES
CREATING MULTIPLE SIZES OF IMAGES
OCR EXTRACTION
EXTRANEOUS OBJECTS NOT PART OF THE CARD, SUCH AS BACKGROUNDS
THE
SOLUTION
We leveraged AI and ML to create an intelligent scanning system that captures high-definition images and automatically crops out all backgrounds while providing a photo output that includes quadrant views of both sides. Additional features include:
CATALOGING BASED ON AUTOMATIC EXTRACTION OF KEY INFORMATION
SORTING OF CARDS BASED ON YEAR, PLAYER, SET, VARIATION, AND CARD NUMBER
OPTIONS TO SEARCH AND FILTER
CREATION OF BEAUTIFUL COLLAGES AND SHOWCASES OF COLLECTED CARDS
SHARING OF COLLECTIONS ON SOCIAL MEDIA
ML Image Processing Flow:
TECH
STACK
THE
RESULTS
IT TAKES ONLY A MINUTE TO CONVERT PHYSICAL CARDS TO DIGITAL COPIES
OVER 50,000 CARDS UPLOADED AND COUNTING
CARD SUPPLY VALUE UPWARDS OF USD 30 MILLION CREATED
NEXTGEM HAS EMERGED AS A COLLECTOR’S PLATFORM OF CHOICE, REALIZING THE FOUNDER’S VISION