为保护服务机器人视觉图像中的敏感个人信息，提出了一种基于离散余弦变换与差分隐私技术相结合的服务机器人图像发布算法。首先，利用离散余弦变换压缩图像，为均衡相关噪声误差和重构误差，引入一种基于随机梯度下降算法的系数选择方法，从而在相应的系数空间中选择出合适的系数来压缩图像；其次，对系数空间添加拉普拉斯噪声来满足ε-差分隐私需求；再次，基于4种真实的室内图像数据集采用小波包变换和最小二乘支持向量机分类技术从算法的查准率等指标上衡量算法。实验结果表明，所提出的图像发布算法有较好的鲁棒性。
To protect the sensitive personal information in service robots, this paper proposes an image publishing algorithm for service robots based on the combination of discrete cosine transform and differential privacy. Firstly, the image is compressed by discrete cosine transform. In order to balance the correlation noise error and reconstruction error, a coefficient selection method based on random gradient descent algorithm is introduced. Then, appropriate coefficients are selected in the corresponding coefficient space to compress the image, and Laplace noise is added to the coefficient space. Finally, based on four real indoor image data sets, wavelet packet transform and least squares support vector machine classification techniques are used to measure the accuracy of the algorithm. The experimental results demonstrate the robustness of the proposed algorithm.
[1] Zhang L, Jung T, Liu K B, et al. PIC:enable large-scale privacy preserving content-based image search on cloud[J]. IEEE Transactions on Parallel and Distributed Systems, 2017, 28(11):3258-3271.
[2] Nakashima Y, Ikeno T, Babaguchi N. Evaluating protection capability for visual privacy information[J]. IEEE Security & Privacy, 2016, 14(1):55-61.
[3] 郭思雨. 基于匿名化技术的人脸图像隐私保护方法研究[D]. 北京:北京交通大学, 2018.
[4] 焦文明. 面向隐私需求的虹膜图像处理方法研究[D]. 徐州:中国矿业大学, 2020.
[5] 孙啸. 结合超声影像区域分类与可逆隐写的医疗隐私保护方法研究[D]. 杭州:浙江财经大学, 2019.
[6] Xia Z H, Wang L, Tang J, et al. A privacy-preserving image retrieval scheme using secure local binary pattern in cloud computing[J]. IEEE Transactions on Network Science and Engineering, 2021, 8(1):318-330.
[7] 熊平, 朱天清, 王晓峰. 差分隐私保护及其应用[J]. 计算机学报, 2014, 37(1):101-122. Xiong P, Zhu T Q, Wang X F. A survey on differential privacy and applications[J]. Chinese Journal of Computers, 2014, 37(1):101-122. (in Chinese)
[8] 张啸剑, 付聪聪, 孟小峰. 面向人脸图像发布的差分隐私保护[J]. 中国图象图形学报, 2018, 23(9):1305-1315 Zhang X J, Fu C C, Meng X F. Facial image publication with differential privacy[J]. Journal of Image and Graphics, 2018, 23(9):1305-1315. (in Chinese)
[9] Dwork C, Kenthapadi K, Mcsherry F, et al. Our data, ourselves:privacy via distributed noise generation[M]//Advances in Cryptology-EUROCRYPT 2006. Berlin, Heidelberg:Springer Berlin Heidelberg, 2006:486-503.
[10] 柯海峰. 基于差分隐私的数据发布技术研究与实现[D]. 南京:南京理工大学, 2018.
[11] Zhang X Q, Tian J X. Fast DNA encoding algorithm inspired by the SPOOLing system[J]. Medical & Biological Engineering & Computing, 2022, 60(9):2707-2720.
[12] Zhang X Q, Liu M, Tian J X, et al. Color image encryption algorithm based on dynamic block zigzag transformation and six-sided star model[J]. Electronics, 2022, 11(16):2512.
[13] Quattoni A, Torralba A. Recognizing indoor scenes[C]//2009 IEEE Conference on Computer Vision and Pattern Recognition, 2009:413-420.
[14] 张啸剑, 付聪聪, 孟小峰. 结合矩阵分解与差分隐私的人脸图像发布[J]. 中国图象图形学报, 2020, 25(4):655-668. Zhang X J, Fu C C, Meng X F. Private facial image publication through matrix decomposition[J]. Journal of Image and Graphics, 2020, 25(4):655-668. (in Chinese)
[15] Han Y K, Choi Y B. Human action recognition based on LSTM model using smartphone sensor[C]//2019 Eleventh International Conference on Ubiquitous and Future Networks (ICUFN), 2019:748-750.
[16] 陶波, 赵兴炜, 李汝鹏, 等. 机器人测量-操作-加工一体化技术研究及其应用[J]. 中国机械工程, 2020, 31(1):49-56. Tao B, Zhao X W, Li R P, et al. Research on robotic measurement-operation-machining technology and its application[J]. China Mechanical Engineering, 2020, 31(1):49-56. (in Chinese)
[17] Yang Y S, Yang C M, Huang C W. Thin crack observation in a reinforced concrete bridge pier test using image processing and analysis[J]. Advances in Engineering Software, 2015, 83(5):99-108.
[18] Chen F, Zhao H, Li D W, et al. Contact force control and vibration suppression in robotic polishing with a smart end effector[J]. Robotics and Computer-Integrated Manufacturing, 2019, 57(6):391-403.
