Theory and Applications of Smart Cameras PDF
Preview Theory and Applications of Smart Cameras
KAIST Research Series Chong-Min Kyung Editor Theory and Applications of Smart Cameras KAIST Research Series Series editors Chan Beum Park, Daejeon, Korea, Republic of (South Korea) Bumki Min, Daejeon, Korea, Republic of (South Korea) Jae Woo Lee, Daejeon, Korea, Republic of (South Korea) Jae Seung Jeong, Daejeon, Korea, Republic of (South Korea) Sang Ouk Kim, Daejeon, Korea, Republic of (South Korea) Insung S. Choi, Daejeon, Korea, Republic of (South Korea) More information about this series at http://www.springer.com/series/11753 Chong-Min Kyung Editor Theory and Applications of Smart Cameras 123 Editor Chong-MinKyung Department ofElectrical Engineering, Centerfor Integrated Smart Sensors KAIST Daejeon Korea,Republic of (SouthKorea) ISSN 2214-2541 ISSN 2214-255X (electronic) KAIST Research Series ISBN978-94-017-9986-7 ISBN978-94-017-9987-4 (eBook) DOI 10.1007/978-94-017-9987-4 LibraryofCongressControlNumber:2015942807 SpringerDordrechtHeidelbergNewYorkLondon ©SpringerScience+BusinessMediaDordrecht2016 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission orinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodologynowknownorhereafterdeveloped. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfrom therelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authorsortheeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinor foranyerrorsoromissionsthatmayhavebeenmade. Printedonacid-freepaper SpringerScience+BusinessMediaB.V.DordrechtispartofSpringerScience+BusinessMedia (www.springer.com) Contents Part I Fundamental/Energy-related Issues of Smart Cameras CMOS Image Sensor for Smart Cameras. . . . . . . . . . . . . . . . . . . . . . 3 JongHo Park Architectural Analysis of a Baseline ISP Pipeline . . . . . . . . . . . . . . . . 21 Hyun Sang Park An Ultra-Low-Power Image Signal Processor for Smart Camera Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Zhenhong Liu and Nam Sung Kim Foundations and Applications of 3D Imaging . . . . . . . . . . . . . . . . . . . 63 Min H. Kim E-R-D Optimization in Video Compression. . . . . . . . . . . . . . . . . . . . . 87 Hyuk-Jae Lee, Hyun Kim and Chae-Eun Rhee Part II Event/Object Detectors for Smart Sensing Low-Power Operation for Video Event Data Recorder . . . . . . . . . . . . 117 Jinyoung Yang, Jongpil Jung and Chong-Min Kyung Low-Power Face Detection for Smart Camera . . . . . . . . . . . . . . . . . . 139 Hyung-Il Kim, Seung Ho Lee and Yong Man Ro Accurate Face and Human Detection Using Hybrid Local Transform Features . . . . . . . . . . . . . . . . . . . . . . 157 Daijin Kim and Bongjin Jun v vi Contents Adaptive Resource Management for Sensor Fusion in Visual Tracking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Bohyung Han, Seong-Wook Joo and Larry S. Davis Traffic Pattern Analysis and Anomaly Detection via Probabilistic Inference Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Hawook Jeong, Youngjoon Yoo, Kwang Moo Yi and Jin Young Choi Event Detection Module for Low-Power Camera. . . . . . . . . . . . . . . . . 241 Byung-geun Lee and Moongu Jeon Advanced Human Detection Using Fused Information of Depth and Intensity Images. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 Gyu-Hong Lee, Dong-Suk Kim and Chong-Min Kyung Part III Wireless Connectivity for Video Sensor Networks Time Synchronization for Multi-hop Surveillance Camera Systems . . . 283 Hyuntae Cho Distributed Medium Access for Camera Sensor Networks: Theory and Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Hojin Lee, Donggyu Yun and Yung Yi Wireless Sensor Network for Video Sensors . . . . . . . . . . . . . . . . . . . . 339 Hyung Won Kim Part I Fundamental/Energy-related Issues of Smart Cameras CMOS Image Sensor for Smart Cameras JongHo Park Abstract A smart camera is a vision system with special features implemented to achieve its specific purpose. A smart camera which can be used for security or surveillance purpose requires high dynamic range of the sensor to cover broad illumination range of the scene. A stick- or badge-type smart camera operates as a stand-alone device so that the power consumption is one of the most important parameters. For applications such as nondestructive inspection using infrared (IR), sensitivity of the image sensors should be improved to obtain suitable SNR for reliableoutput.Thischapterdescribesbasicimagingprinciplesanddynamicrange expansion methods of the CMOS image sensors. (cid:1) (cid:1) Keywords CMOS image sensor (CIS) Charge coupled device (CCD) Active (cid:1) (cid:1) pixel sensor (APS) Wide dynamic range (WDR) Correlated double sampling (CDS) 1 Imaging Principles ImagesensorsinCMOStechnologyareimplementedusinganarrayofsmartphoto sensors called active pixel sensor (APS). The design of the APS is flexible and various types of pixels have been developed for various applications including WDR imaging [1]. The performance of CMOS APS in areas of high-end digital imaging has been proven to be comparable to their CCD counterparts due to the abilityofon-chipimageprocessing[2].Imageprocessingtoacertaindegreecould be performed within the pixel itself by integrating signal processing circuitry in each pixel [3]. Further image processing can be done in the subsequent circuit stages before image information is read out [4]. J.Park(&) CenterforIntegratedSmartSensor,ITCBuilding(N1),KAIST,Daehak-ro291, Yuseong-gu,Deajeon305-701,RepublicofKorea e-mail:[email protected] ©SpringerScience+BusinessMediaDordrecht2016 3 C.Kyung(ed.),TheoryandApplicationsofSmartCameras, KAISTResearchSeries,DOI10.1007/978-94-017-9987-4_1 4 J.Park 1.1 Solid State Imaging Devices 1.1.1 Charge Coupled Device A Charge coupled device (CCD) is the most important technology for image sensors. The CCD offers guaranteed image quality because it uses optimized photodetectors for achieving low noise, low dark current, and high sensitivity [5]. The basic concept of CCD is a simple series connection of MOS capacitors. The individual capacitors are physically located very close to each other, which com- pose an analog shift register driving by two, three or four phase clocks for charge transfer. Figure 1 shows the simplified structure of an interline-transfer CCD. A charge transfer must occur at high enough rates to avoid image corruption by leakage, but slow enough rates to ensure high charge transfer efficiency [5]. The limitation of CCD technology comes mainly from insufficient charge transfer efficiency [6]. To avoid insufficient charge transfer, high-speed and high-voltageclockcontrolschemesinCCDincreasethesystemcomplexityandthe powerconsumption.AnothermajordrawbackinCCDtechnologyisthatperipheral circuits such as ADC cannot be integrated on the same chip. Nevertheless, understanding CCD technology is very important because CMOS image sensors (CISs) have been developed based on CCD. PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD PD V V V V V e e e e e PD rtica PD rtica PD rtica PD rtica PD rtica l l l l l C C C C C C C C C C PD D PD D PD D PD D PD D s s s s s PD PD PD PD PD PD PD PD PD PD Buffer Amp. HorizontalCCDs Charge to Voltage Conversion Fig.1 Simplifiedinterline-transfer(IT)CCDstructure
The list of books you might like
