Low Power Image Acquisition Scheme Using On-Pixel Event Driven Halftoning

Several emerging IOT applications may require ultra-low power image-acquisition techniques, at the cost of relaxed constraints on image quality. Event-driven imaging, based on address-event-representation (AER) proposed earlier, ensures event-driven motion detection and conditional image acquisition. However it relies on delta modulation of pixel data upon event detection, which may lead to error accumulation for grey-scale reconstruction. In this work we propose event-driven on-sensor halftoning scheme based on cellular neural network (CNN), which can achieve compressed image acquisition, reducing the data to one bit per active pixel, while avoiding error accumulation over multiple frames. The proposed scheme can achieve reduction in power dissipation related to image quantization, communication and storage. Event-driven current-mode discrete-time CNN (DT-CNN) circuit is proposed for integration on CMOS imager pixel, which is optimally duty cycled to achieve ultra-low power for the halftoning operation. The impact of design parameters and process variations on the final image quality is assessed through inverse halftoning. Clubbed with event-driven acquisition, the proposed on-sensor halftoning can achieve large reduction in power dissipation and data volume, reduction in total volume, as compared to frame based imaging, as well as event driven grey-scale imaging.

[1]  Tobi Delbrück,et al.  A 128$\times$ 128 120 dB 15 $\mu$s Latency Asynchronous Temporal Contrast Vision Sensor , 2008, IEEE Journal of Solid-State Circuits.

[2]  Bernabé Linares-Barranco,et al.  On Real-Time AER 2-D Convolutions Hardware for Neuromorphic Spike-Based Cortical Processing , 2008, IEEE Transactions on Neural Networks.

[3]  Chang Wen Chen,et al.  Collaborative Image Coding and Transmission over Wireless Sensor Networks , 2007, EURASIP J. Adv. Signal Process..

[4]  Eija Honkavaara,et al.  Point Cloud Generation from Aerial Image Data Acquired by a Quadrocopter Type Micro Unmanned Aerial Vehicle and a Digital Still Camera , 2012, Sensors.

[5]  Tobi Delbrück,et al.  AER Building Blocks for Multi-Layer Multi-Chip Neuromorphic Vision Systems , 2005, NIPS.

[7]  L. Spaanenburg,et al.  Efficient DTCNN implementations for large-neighborhood functions , 1998, 1998 Fifth IEEE International Workshop on Cellular Neural Networks and their Applications. Proceedings (Cat. No.98TH8359).

[8]  Leon O. Chua,et al.  Methods for image processing and pattern formation in Cellular Neural Networks: a tutorial , 1995 .

[9]  Ryad Benosman,et al.  Asynchronous Neuromorphic Event-Driven Image Filtering , 2014, Proceedings of the IEEE.

[10]  Abdullah Al-Dhelaan,et al.  Image-Based Object Identification for Efficient Event-Driven Sensing in Wireless Multimedia Sensor Networks , 2015, Int. J. Distributed Sens. Networks.

[11]  Nigel H. Lovell,et al.  A wearable real-time image processor for a vision prosthesis , 2009, Comput. Methods Programs Biomed..

[12]  Ramesh Govindan,et al.  Image-Based Environmental Monitoring Sensor Application Using an Embedded Wireless Sensor Network , 2014, Sensors.

[13]  Horst Bischof,et al.  Building with drones: Accurate 3D facade reconstruction using MAVs , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[14]  Michael Egmont-Petersen,et al.  Image processing with neural networks - a review , 2002, Pattern Recognit..

[15]  Gert Cauwenberghs,et al.  CMOS Camera With In-Pixel Temporal Change Detection and ADC , 2007, IEEE Journal of Solid-State Circuits.

[16]  Martial Hebert,et al.  Efficient visual event detection using volumetric features , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[17]  Wentai Liu,et al.  An optimal design methodology for inductive power link with class-E amplifier , 2005, IEEE Transactions on Circuits and Systems I: Regular Papers.

[18]  Amine Bermak,et al.  Efficient event-drivern frame capture for CMOS imagers , 2007, 2007 9th International Symposium on Signal Processing and Its Applications.

[19]  Kaushik Roy,et al.  Ultra low energy analog image processing using spin based neurons , 2012, 2012 IEEE/ACM International Symposium on Nanoscale Architectures (NANOARCH).

[20]  Gert Cauwenberghs,et al.  An Active Pixel CMOS separable transform image sensor , 2009, 2009 IEEE International Symposium on Circuits and Systems.