Pre-capture Privacy via Adaptive Single-Pixel Imaging

TL;DR

This paper introduces a pre-capture privacy-preserving imaging method using adaptive single-pixel imaging that eliminates identifiable details of specific targets like faces or license plates while maintaining image utility.

Contribution

It proposes a novel adaptive aperture pattern generator within a single-pixel imaging framework to selectively anonymize targets without hardware changes.

Findings

Effective anonymization of faces and license plates.

Captured images retain utility for computer vision tasks.

Method achieves privacy without sacrificing image quality.

Abstract

As cameras become ubiquitous in our living environment, invasion of privacy is becoming a growing concern. A common approach to privacy preservation is to remove personally identifiable information from a captured image, but there is a risk of the original image being leaked. In this paper, we propose a pre-capture privacy-aware imaging method that captures images from which the details of a pre-specified anonymized target have been eliminated. The proposed method applies a single-pixel imaging framework in which we introduce a feedback mechanism called an aperture pattern generator. The introduced aperture pattern generator adaptively outputs the next aperture pattern to avoid sampling the anonymized target by exploiting the data already acquired as a clue. Furthermore, the anonymized target can be set to any object without changing hardware. Except for detailed features which have been removed from the anonymized target, the captured images are of comparable quality to those captured by a general camera and can be used for various computer vision applications. In our work, we target faces and license plates and experimentally show that the proposed method can capture clear images in which detailed features of the anonymized target are eliminated to achieve both privacy and utility.