How We Detect Fraudulent Images by Beehive ID

BeehiveID provides two types of image matching services, fuzzy content and biometric face.

Fuzzy content matching identifies duplicate images that have been altered to avoid detection. Given an image, our technology will find all exact duplicates of that image, as well as duplicates of that image that have been altered to avoid detection.

Biometric face matching identifies a specific person’s face independent of their facial expression or orientation. All photos of that person — independent of background, lighting, expression or position are identified.

Fuzzy Content

It’s fairly easy to find duplicate images online — there are a lot of free tools. However, the typical methods used do not work if the image is cropped, enlarged or reversed. Fraudsters use simple image processing tools to avoid detection of duplicates. Fuzzy content matching overcomes this limitation and finds images that are similar to each other.

This service is typically used to find duplicate content. Fuzzy matching is generally insensitive to cropping, resizing or compression — it uses the visual similarity of the image for matching.

Biometric Face Matching

Biometric face matching determines whether images of two faces represent the same person. It is commonly used to verify productivity of employees who are working offsite or to prevent cheating for online training or exams. It determines whether a particular individual has maintained a consistent presence at a computer, based up on a series of collected webcam images.

The service can determine both the presence of a face in the webcam image, as well as a change of faces if a different person is now at the computer. It can also detect spoofing of the webcam image by insertion of false images.Visit BeehiveID here.