Image acquisition method and device considering privacy protection and data quality, and storage medium

Abstract

The invention discloses an image acquisition method and device considering privacy protection and data quality, and a storage medium, and aims to solve the technical problem that the image privacy protection effect and the data quality are difficult to consider at the same time. The method comprises the following steps: acquiring a plurality of original images according to an image collection task; carrying out differential privacy protection and encoding on the original images to obtain noise-added images and hash codes of the noise-added images; carrying out iterative clustering on the hashcodes of all the noise-added images to form N clusters, and each cluster comprises the hash codes; performing hash code screening according to the Hamming distance between each hash code in a clusterand the centroid point of the cluster to obtain high-quality hash code; and extracting a high-quality image from all the noise-added images according to the high-quality hash code, and taking the high-quality image as a final image acquisition result. Image privacy protection and data quality can be well considered, and the images not prone to privacy leakage, high in availability, high in qualityand high in reliability is obtained.

Description

technical field [0001] The invention relates to an image collection method, device and storage medium that take into account both privacy protection and data quality, and belongs to the technical field of data collection. Background technique [0002] Images have now become an important source of information for enterprises and scientific researchers, especially image datasets with large quantities, high availability, and reliable quality. By mining and analyzing the obtained image data sets, image collectors or users can provide more personalized services for enterprises and provide richer data support for researchers. However, digital images usually contain a wealth of private information. If these images are released directly, the sensitive information in them will be leaked, which will become a way for attackers to benefit. It can accurately locate the user's location; the personal image posted by the user on the network will be modified by the attacker to use it as fra...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the image privacy protection effect and data quality are difficult to balance in the prior art, the present invention proposes an image collection method, device and storage medium that takes into account both privacy protection and data quality, compresses and transforms the image, and selectively adds Noise, to achieve a balance between image privacy protection and usability, and use image coding to perform high-quality screening of noisy images to ensure the data quality of collected images, and then take into account the effect of image privacy protection and data quality

Images