Search

Privacy awareness constitutes a critical aspect for smart camera networks. An ideal awless protection of sensitive information would boost their application scenarios. How- ever, it is still far from being achieved. Numerous chal- lenges arise at di erent levels, from hardware security to subjective perception. Generally speaking, it can be stated that the closer to the image sensing device the protection measures take place, the higher the privacy and security at- tainable. Likewise, the integration of heterogeneous camera components becomes simpler since most of them will not require to consider privacy issues. The ultimate objective would be to incorporate complete protection directly into a smart image sensor in such a way that no sensitive data would be delivered o -chip while still permitting the tar- geted video analytics. This paper presents a 320 240-px prototype vision sensor embedding processing capabilities useful for accomplishing this objective. It is based on re- con gurable focal-plane sensing-processing that can provide programmable obfuscation. Pixelation of tunable granular- ity can be applied to multiple image regions in parallel. In addition to this functionality, the sensor exploits recon g- urability to implement other processing primitives, namely block-wise high dynamic range, integral image computation and Gaussian ltering. Its power consumption ranges from 42.6mW for high dynamic range operation to 55.2mW for integral image computation at 30fps. It has been fabricated in a standard 0.18 m CMOS process.

Privacy concerns are hindering the introduction of smart camera networks in prospective application scenarios like retail analytics, factory monitoring or elderly care. The idea of networked cameras pervasively collecting data gen- erates social rejection in the face of sensitive information being tampered by hackers or misused by legitimate users. New strategies must be developed in order to ensure pri- vacy from the very point where sensitive data are generated: the sensors. Protection measures embedded on-chip at the front-end sensor of each network node signi cantly reduce the number of trusted system components as well as the im- pact of potential software aws. In this demonstration, we present a full-custom QVGA vision sensor that can be re- con gured to implement programmable pixelation of image regions at the focal plane. In particular, we show on-the- y focal-plane face obfuscation supported by the Viola-Jones frontal face detector provided by OpenCV

Privacy concerns are hindering the introduction of smart camera networks in prospective application scenarios like retail analytics, factory monitoring or elderly care. The idea of networked cameras pervasively collecting data gen- erates social rejection in the face of sensitive information being tampered by hackers or misused by legitimate users. New strategies must be developed in order to ensure pri- vacy from the very point where sensitive data are generated: the sensors. Protection measures embedded on-chip at the front-end sensor of each network node signi cantly reduce the number of trusted system components as well as the im- pact of potential software aws. In this demonstration, we present a full-custom QVGA vision sensor that can be re- con gured to implement programmable pixelation of image regions at the focal plane. In particular, we show on-the- y focal-plane face obfuscation supported by the Viola-Jones frontal face detector provided by OpenCV

Privacy awareness constitutes a critical aspect for smart camera networks. An ideal awless protection of sensitive information would boost their application scenarios. How- ever, it is still far from being achieved. Numerous chal- lenges arise at di erent levels, from hardware security to subjective perception. Generally speaking, it can be stated that the closer to the image sensing device the protection measures take place, the higher the privacy and security at- tainable. Likewise, the integration of heterogeneous camera components becomes simpler since most of them will not require to consider privacy issues. The ultimate objective would be to incorporate complete protection directly into a smart image sensor in such a way that no sensitive data would be delivered o -chip while still permitting the tar- geted video analytics. This paper presents a 320 240-px prototype vision sensor embedding processing capabilities useful for accomplishing this objective. It is based on re- con gurable focal-plane sensing-processing that can provide programmable obfuscation. Pixelation of tunable granular- ity can be applied to multiple image regions in parallel. In addition to this functionality, the sensor exploits recon g- urability to implement other processing primitives, namely block-wise high dynamic range, integral image computation and Gaussian ltering. Its power consumption ranges from 42.6mW for high dynamic range operation to 55.2mW for integral image computation at 30fps. It has been fabricated in a standard 0.18 m CMOS process.