本論文提出一個多模式動態偵測影像感測器，可提供8-bit灰階影像、2-bit事件偵測和8-bit幀間差異的暫態差異感測、以及畫素陣列子區塊中動態事件累計的顯著偵測。所提出的在乒乓¬-脈衝寬度調節器(Pulse Width Modulation)設計於畫素電路中以完成相鄰兩幀的相減來實現連續圖像視覺系統上的動態偵測。
為了在畫素電路中達到相鄰兩幀訊號相減的功能，使用MIMCAP電容作為儲存位元並以電荷耦合操作讓兩個訊號能從電荷面相減，並且結合兩組相同電路成乒乓架構來交錯儲存模式和相減讀出模式，因此可達到連續相減訊號讀出以維持資料傳輸率。畫素中脈衝寬度調節器提供低電壓操作和訊號數位化讀出，脈衝寬度訊號傳至縱向共用的數位計數器讀出，使整體架構功耗降低且傳輸速率提升。在動態追蹤的幀間差異感測和顯著偵測模式下，感光操作採用全局式快門來避免動態模糊。顯著偵測模式統計畫素陣列8×8區塊中動態事件的數量，利用同時讀取八行的電流總量並由縱向共用的電容反饋跨阻抗放大器(Capacitive Trans-Impedance Amplifier)和單斜坡類比數位轉換器(Single-Slope Analog-to-Digital Converter)回報區塊中事件的總數。
為了驗證本電路，一個操作於0.8V並搭載64×64乒乓PWM畫素陣列的多模式動態偵測影像感測器原型使用0.18微米1P6M互補式金氧半導體製作，晶片總面積為2000×2100平方微米。量測結果顯示此晶片在影像擷取模式下達到每秒360幀、功耗71.2微瓦；在8-bit幀間差異模式下達到每秒510幀、74.4微瓦；以及區塊顯著偵測模式下每秒890幀、121.6微瓦。

This thesis presents a multi-mode motion detection CMOS image sensor, providing 8-bit gray level image capturing(IC), temporal difference reporting with 8-bit frame difference(FD) and 2-bit event detection(ED), and saliency detection(SD) with block-level event counting. The proposed ping-pong pulse width modulation(PWM) is implemented in pixel circuit for consecutive frames subtraction to realize motion detection for frame-based vision system
To achieve frame difference between two consecutive frames in pixel circuit, MIMCAP capacitor is implemented as memory cell and used for subtraction by charge domain switching capacity technique. By combining two identical circuits into ping-pong structure, two memory cells are operated in storage and differencing phase separately so that consecutive difference signal can be achieved, maintaining original data rate. In pixel PWM supports low voltage operation and time domain digitization readout, transferring pulse width signal to column-shared digital counter, for lower power consumption and higher data bandwidth. For motion detection FD and SD modes, global shutter is adopted to avoid motion blur. In SD mode, motion events in every 8×8 blocks of pixel array are estimated by turning on 8 rows simultaneously. Column-shared capacitive trans-impedance amplifier(CTIA) sums up 8 event currents, and Single-slope analog-to-digital converter(SS-ADC) converts output into digital code.
An 0.8V global shutter multi-mode motion detection vision sensor prototype with 64×64 ping-pong PWM pixel array has been fabricated in 0.18μm 1P6M standard CMOS technology with a chip area of 2000×2100μm2. The measurement result shows that the proposed sensor achieves 360 fps and 71.2μW in IC mode, 510 fps and 74.4μW in FD mode, 890 fps and 121.6μW in SD mode.

ABSTRACT II
CONTENTS III
LIST OF FIGURES VI
LIST OF TABLES IX
CHAPTER 1 INTRODUCTION 1
1.1 MOTIVATION 1
1.2 THESIS CONTRIBUTION 2
1.3 THESIS ORGANIZATION 4
CHAPTER 2 BACKGROUND INFORMATION 6
2.1 FUNDAMENTALS OF CMOS IMAGE SENSOR 7
2.1.1 Active Pixel Sensor(APS) 7
2.1.1.1 3T Active Pixel Sensor 8
2.1.1.2 4T Active Pixel Sensor 9
2.1.1.3 PWM Pixel Sensor 11
2.1.1.4 PFM Pixel Sensor 12
2.1.2 Pixel Readout 13
2.1.2.1 Rolling Shutter 13
2.1.2.2 Global Shutter 14
2.1.3 Fundamental Terms of Image Sensor 15
2.1.3.1 Pixel Pitch 16
2.1.3.2 Fill Factor 16
2.1.3.3 Exposure Time 16
2.1.3.4 Frame Rate 17
2.1.3.5 Lux Transfer Curve 17
2.2 DYNAMIC VISION SENSOR(DVS) 17
2.2.1 Event-based DVS 18
2.2.2 Frame-based DVS 21
2.3 MOTION SALIENCY ALGORITHM 23
2.4 SUMMARY 25
CHAPTER 3 PROPOSED FD PING-PONG PWM PIXEL 27
3.1 TVC PWM PIXEL 28
3.2 CHARGE-DOMAIN SUBTRACTION PWM PIXEL 30
3.3 PING-PONG PWM PIXEL 32
3.4 SUMMARY 34
CHAPTER 4 CIRCUIT IMPLEMENTATION 35
4.1 SYSTEM ARCHITECTURE 35
4.1.1 Pixel Circuit 37
4.1.2 Vth & Ramp Generator 38
4.1.3 Column Shared Circuit 38
4.2 OPERATION MODE 40
4.2.1 Image Capture(IC) Mode 41
4.2.2 Frame Difference(FD) Mode 43
4.2.3 Saliency Detection(SD) Mode 44
4.3 SUMMARY 46
CHAPTER 5 MEASUREMENT RESULTS 47
5.1 CHIP IMPLEMENTATION 47
5.2 MEASUREMENT ENVIRONMENT SETUP 49
5.3 SENSOR MEASUREMENT 50
5.3.1 IC Mode 50
5.3.2 FD Mode 52
5.3.3 SD Mode 55
5.4 PERFORMANCE SUMMARY AND COMPARISON 56
CHAPTER 6 CONCLUSION AND FUTURE WORK 58
6.1 CONCLUSION 58
6.2 FUTURE WORK 59
BIBLIOGRAPHY 61

[1] J. Ohta, "Smart CMOS Image Sensor and Applications," CRC Press.
[2] J. Nakamura, "Image Sensor and Signal Processing for Digital Still Camera, Taylor & Francis Group, 2006.
[3] P. Lichtsteiner and T. Delbruck et al., "A 128 128 120 dB 15 μs Latency Asynchronous Temporal Contrast Vision Sensor," IEEE Journal of Solid-State Circuits, vol. 43, no. 2, February 2008.
[4] J. A. Leñero-Bardallo et al., "A 3.6 s Latency Asynchronous Frame-Free Event-Driven Dynamic-Vision-Sensor, " IEEE Journal of Solid-State Circuits, vol. 46, no. 6, June 2011.
[5] T. Serrano-Gotarredona and B. Linares-Barranco, "A 128 128 1.5% Contrast Sensitivity 0.9% FPN 3 μs Latency 4 mW Asynchronous Frame-Free Dynamic Vision Sensor Using Transimpedance Preamplifiers," IEEE Journal of Solid-State Circuits, vol. 48, no. 3, June 2013.
[6] C. Brandli and T. Delbruck et al., "A 240 × 180 130 dB 3 μs Latency Global Shutter Spatiotemporal Vision Sensor," IEEE Journal of Solid-State Circuits, vol. 49, no. 10, October 2014.
[7] M. Yang and T. Delbruck et al., "A Dynamic Vision Sensor With 1% Temporal Contrast Sensitivity and In-Pixel Asynchronous Delta Modulator for Event Encoding, " IEEE Journal of Solid-State Circuits, vol. 50, no. 9, September 2015.
[8] B. Son et al., "A 640×480 Dynamic Vision Sensor with a 9μm Pixel and 300Meps Address-Event Representation," IEEE ISSCC Dig. Tech. Papers, pp. 66-67, Feb. 2017.
[9] C. Li, et al., "A 132 by 104 10μm-Pixel 250μW 1kefps Dynamic Vision Sensor with Pixel-Parallel Noise and Spatial Redundancy Suppression," IEEE Symp. VLSI Circuits, pp. 216-217, June 2019.
[10] R. Benosman et al., "Asynchronous Frameless Event-based Optical Flow," Neural Networks, vol. 27, pp 32-37, March 2012.
[11] R. Benosman et al., "Event-Based Visual Flow," IEEE Transactions on Neural Network and Learning System, vol. 25, no. 2, February 2014.
[12] A.Borst, et al., "Fly Motion Vision," Annual Review of Neuroscience, vol. 33, pp 49-70, March 2010.
[13] L. Luo, et al., "Video Saliency Detection Using Motion Saliency Filter," Proceeding of 2013 3rd International Conference on Computer Science and Network Technology, pp.1045-1049, Oct. 2013.
[14] S. Park and E. Yoon, "243.3pJ/pixel Bio-Inspired Time-Stamp-Based 2D Optic Flow Sensor for Artificial Compound Eyes," IEEE ISSCC Dig. Tech. Papers, pp. 126-127, Feb. 2014.
[15] S. Shishido et al., "210ke-Saturation Signal 3μm-Pixel Variable-Sensitivity Global-Shutter Organic Photoconductive Image Sensor for Motion Capture," IEEE ISSCC Dig. Tech. Papers, pp. 122-113, Feb. 2016.
[16] K. D. Choo et al., "Energy-Efficient Low-Noise CMOS Image Sensor with Capacitor Array-Assisted Charge-Injection SAR ADC for Motion-Triggered Low-Power IoT Applications," IEEE ISSCC Dig. Tech. Papers, pp. 96-98, Feb. 2019.
[17] Berthold K. P. Horn and Brian G. Schunck, "Determining Optical Flow," Artificial Intelligence, vol. 17, issue1-3, pp 185-203, August 1981.
[18] J. Choi, S. Park, J. Chao and E. Yoon, "A 3.4-μW Object-Adaptive CMOS Image Sensor with Embedded Feature Extraction Algorithm for Motion-Triggered Object-of-Interest Imaging," IEEE Journal of Solid-State Circuits, vol. 49, no. 1, January 2014.
[19] T. Ohmaru et al., “A 25.3μW at 60 fps 240×160 Pixel Vision Sensor for Motion Capturing With In-Pixel Nonvolatile Analog Memory Using CAAC-IGZO FET ,“ IEEE Journal of Solid-State Circuits, vol.51, no. 9, September 2016.
[20] G. Kim, et al., “A 467nW CMOS Visual Motion Sensor with Temporal Averaging and Pixel Aggregation,” IEEE ISSCC Dig. Tech. Papers, pp. 480-481, Feb. 2013.
[21] A. Y. Chiou, et al., “An ULV PWM CMOS Imager With Adaptive-Multiple-Sampling Linear Response, HDR Imaging, and Energy Harvesting,” IEEE J. Solid-State Circuits, vol. 54, no. 1, pp. 298-306, Jan. 2019.
[22] C. Yin, et al., “A 0.5 V, 14.28 kfps, 109 dB Smart Image Sensor With Array-Level Image Signal Processing for IoT Applications,” IEEE Transaction on Electron Devices, vol. 63, no. 3, pp. 1134-1140, Mar. 2016.
[23] C. Posch, et al., “QVGA 143 dB Dynamic Range Frame-Free PWM Image Sensor with Lossless Pixel-Level Video Compression and Time-Domain CDS,” IEEE J. Solid-State Circuits, vol. 46, no. 1, pp. 259-275, Jan 2011
[24] K. Lee, et al., “A 272.49pJ/pixel CMOS Image Sensor with Embedded Object Detection and Bio-Inspired 2D Optic Flow Generation for Nano-Air-Vehicle Navigation,” IEEE Symp. VLSI Circuits, pp. 294-295, June 2017.