Without having read any more than can be gleaned from the article, it sounds like it's using similar principles. That is, the image is transformed into a domain

Without having read any more than can be gleaned from the article, it sounds like it's using similar principles. That is, the image is transformed into a domain (frequency) that is more amenable to rapidly capturing the necessary information quickly without time gating directly. It's not obvious to me what the setup is, but this type of approach is neat.