汽车安全系统中的摄像头技术
在上世纪60年代，汽车业将注意力转向寻找能够保证驾驶者和乘客安全的创新方法。结果如何呢？出现了缓冲方向盘和仪表板、座椅安全带，以及其后的安全气囊和电子稳定性控制系统。自那时起，汽车工程师便一直在探索保护行人、驾驶者和乘客安全的方法。因此，汽车业在提高安全性方面取得了巨大进步，从制动防抱死和多个安全气囊，以至后视摄像系统和先进的导航系统。
In the 1960s, the automotive industry turned its attention to finding creative ways to keep drivers and passengers safe. The result? Padded steering wheels and dashboards, seat belts, and later, airbags and electronic stability control. Since that time, automotive engineers have continued to explore ways to keep pedestrians, drivers and vehicle passengers safe. As a result, the industry has seen tremendous growth in the deployment of safety enhancements — from anti-lock brakes and multiple airbags per vehicle to rearview camera systems and advanced navigation systems.
如今，汽车业继续受到来自政府、非政府组织以及消费者的压力，以期制造行驶更安全的车辆。例如，美国国家公路交通安全管理局(NHTSA)提出，必须明确规定驾驶员在倒车时能够看见的车后方区域。预计在美国销售的乘用车将会全部采用这一提议，以便大幅减少由倒车事故造成的灾祸和伤亡，涉及儿童、残疾人、老年人以及其它行人。
Today, the automotive industry is under continued pressure from governmental bodies, non-governmental organizations and consumer themselves to produce vehicles that are safer to operate. The National Highway Transportation Safety Administration NHTSA, for example, has proposed that a specific area immediately behind each vehicle be defined in which the driver must be able to see when the vehicle’s transmission is in reverse. Adoption of this proposal for all passenger vehicles sold in the U.S. is projected to significantly reduce fatalities and injuries caused by backover accidents involving children, persons with disabilities, the elderly, and other pedestrians.
因此，基于摄像头的视觉系统便成为实施这项提议的关键所在。美国国家公路交通安全管理局要求：2014年9月1日及其以后在美国制造的所有车辆必须带有后视系统。这一要求推动了摄像头解决方案的生产量增长，推动业界开发性能更高、分辨率更好的成像解决方案，并且推动了下一代安全应用和安全系统的发展。
Camera-based visibility systems are therefore critical to the execution and deployment of this proposal. The NHTSA’s proposal, which requires that all vehicles manufactured on or after September 1, 2014 in the U.S. include rear visibility systems, will drive the production volume of camera solutions, the development of higher performance, higher resolution imaging solutions, and the evolution of next-generation safety applications and systems.
面向新兴汽车市场的摄像头
Cameras for an emerging automotive market
在汽车行业，摄像头逐渐成为高级驾驶员辅助系统(ADDAS)的重要组成部分。这些摄像头用于多种用途，例如：交通标志识别、用于停车距离控制(PDC)和车道偏离警示系统(LDW)的后视系统。如今，许多新型车辆集成有数个摄像头，覆盖了汽车周围的所有区域，因而能够实现最优化的安全系统。
Increasingly, cameras are becoming a critical part of advanced driver assistance solutions (ADAS) for the automotive industry. These cameras can be used for multiple applications, such as road sign recognition, rear view for park distance control (PDC) and lane departure warning (LDW). Today, many new model vehicles have several cameras integrated and positioned to cover the entire surround of the car, thereby enabling optimized safety systems.
在这些先进的汽车摄像头安全系统中，增强的成像和感测能力开启了一个无限可能的世界。摄像头视觉系统采用远视场来补足长距离雷达，使用较宽视场来弥补短距雷达或超声波传感器的不足，使得驾驶员可以看到前后方的道路。因此，可提供另一组数据来确认所绘制的整体图片的精确性。例如，来自雷达和摄像头的支持数据能够大大增强决策的信心，如使用刹车系统等等。
Within these advanced automotive camera-based safety systems, enhanced imaging and sensing has opened up a world of possibilities. Camera based vision systems complement long range radar with a long field of view and short range radar or ultrasonic sensor with a wider field of view, allowing the driver to see the road ahead and behind. As a result, another set of data becomes available to confirm the accuracy of the overall picture being drawn. For example, supporting data from radar and cameras provides a significantly higher level of confidence in decision making, such as applying the braking system.
汽车摄像头安全系统内的图像传感器
Image sensors within automotive camera-based safety systems
自1969年电荷耦合器件(CCD)出现以来，其应用一直在增长，扩展到数码相机、望远镜、摄像机、扫描仪，以及其它成像应用领域。然而，CCD在汽车安全应用方面有其固有不足之处。最明显的是：CCD制造需要专门的工厂、设备及工艺，这给以低成本大批量生产汽车摄像头带来了困难。在汽车安全方面，CCD架构也有内在的局限性。标准CCD图像传感器经设计以串行方式读出数据，这意味着必须读出所有的先前像素之后，才能读出后面的像素。这限制了帧速，或每秒读取的图像数目，在快速捕捉图像为关键性功能的汽车安全应用中成为一种重要的缺陷。此外，大多数CCD摄像头不具备表1突显的必需的宽动态范围(WDR)功能。
Since their inception in 1969, charge coupled devices (CCDs) have grown in use. CCDs expanded into digital cameras, telescopes, video cameras, scanners, and other imaging applications. However, CCDs have inherent draw-backs for automotive safety. Most notably, CCD manufacturing requires specialized foundries, equipment, and processes, which prevents them from meeting the low-cost, high-volume production requirements for automotive cameras. The CCD architecture also poses inherent limitations for automotive safety. Standard CCD image sensors are designed for serial readout, meaning that all previous pixels must be read out before the next pixels can be read. Because this limits the frame rate, or number of images read per second, this is a significant disadvantage in automotive safety applications, where capturing images rapidly is a critical function. Additionally, most CCD-based cameras do not provide the necessary wide dynamic range (WDR) as highlighted in Table 1.
表1：汽车摄像头的要求
汽车摄像头要求 描述 传感器要求
高成像灵敏度
或低光照性能 较高的光转换成信号效率，不产生额外的噪声。更高的灵敏度意味着实现具有优良信噪比(SNR)和可用图像所需的光较少。具有高灵敏度的摄像头可以低光照(<0.01 lux)条件下捕捉到良好的图像 需要具备最佳的低光照性能、宽分辨率范围的VGA SOC、WVGA和百万像素传感器。
宽动态范围 在照明条件极端变化下具有高性能。能够在接近前灯、来自其它车辆的眩光、进出隧道、日出和日落等条件下精确捕捉视觉信息，能够缩短曝光响应时间。 具有120dB动态范围的WVGA和百万像素传感器
包括近红外的宽光谱范围灵敏度 近红外(NIR)波长灵敏度可让摄像头捕捉到夜空中出现的波长。也可以通过特殊的前灯来投射近红外光，作为辅助照明，因为人眼是看不到的。 可在白天和夜晚条件下工作
单色和彩色处理
宽光谱范围：400nm到1100nm
近红外灵敏度
高可靠性并符合汽车质量标准 摄像头必须能够耐受汽车使用环境，应当按照ISO/TS16949汽车零部件国际质量管理体系来生产和交货。必须保持高水平的产品质量 生产过程中的严格质量控制
零缺陷
低安全系统总体成本 汽车摄像头统的价格必须对消费者具有吸引力，这是通过控制摄像头自身元件的成本，以及通过摄像头设计来降低总体系统成本的方式来实现 适应性：适合多种应用编程的可配置能力；系统级芯片(SoC)灵活性；降低其它系统元件成本的先进功能
互补金属氧化物半导体(CMOS)图像传感器是替代的解决方案，它可在具有高成本效益的商业代工厂生产，使用与生产其它计算芯片相同的高量产工艺。CMOS图像传感器具有随机访问读取功能，这意味着可对像素进行随机寻址，并可快速读出子帧(sub-frame)，即感兴趣的区域。因此，CMOS图像传感器克服了上述CCD传感器的某些缺点，而且具有几乎无限的子采样和子窗口能力。CMOS摄像头的一个附加优势是能够满足宽动态范围(WDR)和低光照性能要求，进一步推动CMOS图像传感器取代CCD传感器而成为众多成像市场(包括汽车市场)的首选传感器(见图1)。实际上，CMOS汽车传感器市场可望从2009年的不到1000万个增长至2017年的5000万个以上(IMS 2009年汽车市场报告)。
Alternatively, complementary metal oxide semiconductor (CMOS) image sensors are designed for fabrication in cost-effective commercial foundries, using the same high-yielding processes that produce other computing chips. CMOS image sensors have random access readout, meaning that pixels can be randomly addressed and a sub-frame (region of interest) quickly read out. As a result, CMOS image sensors overcome some of the disadvantages of CCDs described above and are nearly unlimited in their ability to sub sample and sub window. As an additional benefit, CMOS cameras are capable of meeting the requirements for WDR and low-light performance, further driving the CMOS image sensors to replace CCDs as the image sensors of choice in numerous imaging markets, including automotive (see automotive application use in Figure 1). In fact, the market for CMOS automotive sensors is forecasted to grow from less than 10 million units in 2009 to more than 50 million units by 2017 (IMS 2009 automotive market report).
在拥有摄像头安全系统车辆内，使用图像传感器来完成场景观察或场景处理。
Within the vehicle’s camera-based safety systems, image sensors are utilized for scene-viewing applications or scene-processing applications.
------------------------------------------
图1 汽车应用中的CMOS图像传感器
场景观察
Scene viewing
通常在视频显示器上输出的场景观察应用包括后视摄像头、多摄像头360度视野、辅助停车、替代后视镜、夜视和后排乘客监视。在这些应用中，关键的图像传感器要求包括出色的低光照性能、WDR和低功耗工作，同时满足汽车质量标准要求，在大批量生产下具有低的单位成本。
Scene-viewing applications which typically output on a video display, include rear view camera, multi-camera 360 degree view, parking assist, wing mirror replacement, night vision and back seat passenger monitoring. In these applications, key image sensor requirements include excellent low-light performance, WDR and low power operation while meeting automotive quality standards and providing low unit costs at high volume.
Aptina imaging公司即是CMOS图像传感器供应商的典范，面向场景观察应用提供全集成式成像解决方案。
例如：专为汽车场景观察系统而设计的VGA系统级芯片(MT9V128和MT9V129)，以及用于下一代视觉摄像头的先进百万像素(MT9M024)解决方案。
Aptina imaging is an example of a CMOS image sensor providor targting scene viewing applicaitons with fully-integrated imaging solutions, such as VGA system on a chip (SOC‘s) designed specifically for automotive scene viewing systems (MT9V128 & MT9V129) and of advanced megapixel (MT9M024) solutions for next generation of viewing cameras.
场景处理
Scene processing
在场景处理应用中，处理单元使用成像器的输出来发送警报或做出有关车辆运行的决策，包括车道偏离警示、交通标志检测、下雨检测、前灯调暗和盲点检测。
Scene-processing applications, where the output of the imager is used by the processing unit to send alerts or make decisions regarding vehicle operation, include lane departure warning, road sign detection, rain detection, headlight dimming and blind spot detection.
使用这些智能化安全系统，处理计算机接收来自传感器的数据，做出决定，并向车辆子系统发送指令，以防止发生事故、减轻事故的严重程度或者保护车主。例如处理计算机可能使用来自图像传感器的数据，确定即将发生的碰撞，并向车辆子系统发出指令以采取刹车措施。
With these intelligent safety system applications, the processing computer receives data from the sensor as input, makes a decision, and sends commands to vehicle subsystems to help prevent an accident, mitigate its severity, or protect the vehicle occupants. For example, a processing computer might use the data from the image sensor to decide that a collision is imminent and send a command to the vehicle subsystems to apply brakes.
微处理器和图像传感器方面的技术进步使得智能化安全系统成为现实。随着技术进步的速度加快，集成更多智能性的图像传感器将成为被动和主动型安全系统的组成部分。
Technological advances in microprocessors and image sensor technologies are making intelligent safety systems a reality. As these advances accelerate, image sensors with integrated intelligence will become part of both passive and active safety systems.
汽车安全性的未来支持
What the future holds in automotive safety
新兴的汽车摄像头安全应用正在以惊人的速度增长，这些应用的配售率一直在上升。地区性标准、交通状况以及驾驶者偏好决定了可用的功能。例如在美国销售的汽车具有后视摄像头、车道偏离警示和360度环绕视觉功能，而欧洲的驾驶者除了360度环绕视觉摄像头外，还能够享有后视摄像头、车道偏离警示、自动远光控制、夜视和交通标志识别功能，每辆车需要多达八个或者更多的摄像头。
Emerging automotive camera-based safety applications are growing at astounding pace and the attach rate for these applications has been on the rise. Regional standards, traffic conditions and driver preferences dictate the features available. For example, cars sold in the U.S. have rear view camera, lane departure warning and or 360 degree surround-view offerings while European drivers see rear view camera, lane departure warning, auto high beam control, night vision and traffic sign recognition features offered in addition to 360 degree surround-view cameras, requiring as many as eight or more cameras per vehicle.
图2 多摄像头360度/环绕视觉系统拓扑
由于每辆车的功能和摄像头数量都在增加，提供集成新功能和新接口的灵活性，对于图像传感器的制造商而言是十分重要的。无论是在摄像头模块内采用协同芯片方式，或者在电子控制单元(ECU)内，最好的办法是使传感器和图像传感器处理ISP功能分开。
As the number of features and the number of cameras per vehicle grow, it is important for image sensor manufacturers to provide the flexibility for the integration of new functions and interfaces. The best approach for achieving this is to separate sensor and Image Sensor Processing ISP function, either within the camera module through a companion chip or in the Electronic Control Unit ECU.
图像传感器制造厂商正在努力提高分辨率，从VGA到百万像素。更高的分辨率可以获得更清晰的街角和边缘。当然也要控制系统的功耗和成本，最大限度地降低摄像头模块的功耗，同时可以获得最佳的图像质量。一种降低成本的方法是使用可用于场景观察和场景处理两种应用的传感器。这就要求根据系统结构、接口和应用状况，在主要针对显示的场景观察和发生在ECU中的场景处理之间复用传感器的输出。Aptina imaging公司开发了解决这些难题的方法，通过提供能够用于摄像头模块和ECU的百万像素传感器以及协同芯片的解决方案，同时实现了更低的功耗和出色的灵活性。
Manufacturers are also working to increase resolution from VGA to megapixel. Higher resolution allows for better definition of corners and edges. Of course, there is also a need to keep power consumption and cost of the system in check. Minimizing power dissipation in the camera module allows for the highest image quality. One way to reduce cost would be to implement sensors that can be utilized in both scene-viewing and scene-processing applications. This requires sensor output to be multiplexed between viewing directed primarily at the display and processing occurring in the ECU, depending on the system architecture, interface and application. An approach to solving these challenges being advanced by Aptina achieves both lower power and excellent flexibility by providing megapixel sensor + companion chip solutions that can be deployed in the camera module and on in the ECU.
结论
Conclusion
汽车业在生产行驶更安全的车辆方面继续承受越来越大的压力。如今，摄像头视觉系统已成为开发更先进安全系统的关键。在这些系统中，CMOS图像传感器不仅能够满足WDR和低光照性能要求，还可提供对于价格敏感的汽车市场至关重要的规模经济效益，有助于该领域实现显著增长。
More and more, the automotive industry is under continued pressure to produce vehicles that are safer to operate. Today, camera-based visibility systems are critical to the development of more advanced safety systems. Within these systems, CMOS image sensors are not only meeting the requirements for WDR and low-light performance, they are also providing the economies of scale critical to the price-sensitive automotive market, contributing to significant growth in the segment.
In the 1960s, the automotive industry turned its attention to finding creative ways to keep drivers and passengers safe. The result? Padded steering wheels and dashboards, seat belts, and later, airbags and electronic stability control. Since that time, automotive engineers have continued to explore ways to keep pedestrians, drivers and vehicle passengers safe. As a result, the industry has seen tremendous growth in the deployment of safety enhancements — from anti-lock brakes and multiple airbags per vehicle to rearview camera systems and advanced navigation systems.
如今，汽车业继续受到来自政府、非政府组织以及消费者的压力，以期制造行驶更安全的车辆。例如，美国国家公路交通安全管理局(NHTSA)提出，必须明确规定驾驶员在倒车时能够看见的车后方区域。预计在美国销售的乘用车将会全部采用这一提议，以便大幅减少由倒车事故造成的灾祸和伤亡，涉及儿童、残疾人、老年人以及其它行人。
Today, the automotive industry is under continued pressure from governmental bodies, non-governmental organizations and consumer themselves to produce vehicles that are safer to operate. The National Highway Transportation Safety Administration NHTSA, for example, has proposed that a specific area immediately behind each vehicle be defined in which the driver must be able to see when the vehicle’s transmission is in reverse. Adoption of this proposal for all passenger vehicles sold in the U.S. is projected to significantly reduce fatalities and injuries caused by backover accidents involving children, persons with disabilities, the elderly, and other pedestrians.
因此，基于摄像头的视觉系统便成为实施这项提议的关键所在。美国国家公路交通安全管理局要求：2014年9月1日及其以后在美国制造的所有车辆必须带有后视系统。这一要求推动了摄像头解决方案的生产量增长，推动业界开发性能更高、分辨率更好的成像解决方案，并且推动了下一代安全应用和安全系统的发展。
Camera-based visibility systems are therefore critical to the execution and deployment of this proposal. The NHTSA’s proposal, which requires that all vehicles manufactured on or after September 1, 2014 in the U.S. include rear visibility systems, will drive the production volume of camera solutions, the development of higher performance, higher resolution imaging solutions, and the evolution of next-generation safety applications and systems.
面向新兴汽车市场的摄像头
Cameras for an emerging automotive market
在汽车行业，摄像头逐渐成为高级驾驶员辅助系统(ADDAS)的重要组成部分。这些摄像头用于多种用途，例如：交通标志识别、用于停车距离控制(PDC)和车道偏离警示系统(LDW)的后视系统。如今，许多新型车辆集成有数个摄像头，覆盖了汽车周围的所有区域，因而能够实现最优化的安全系统。
Increasingly, cameras are becoming a critical part of advanced driver assistance solutions (ADAS) for the automotive industry. These cameras can be used for multiple applications, such as road sign recognition, rear view for park distance control (PDC) and lane departure warning (LDW). Today, many new model vehicles have several cameras integrated and positioned to cover the entire surround of the car, thereby enabling optimized safety systems.
在这些先进的汽车摄像头安全系统中，增强的成像和感测能力开启了一个无限可能的世界。摄像头视觉系统采用远视场来补足长距离雷达，使用较宽视场来弥补短距雷达或超声波传感器的不足，使得驾驶员可以看到前后方的道路。因此，可提供另一组数据来确认所绘制的整体图片的精确性。例如，来自雷达和摄像头的支持数据能够大大增强决策的信心，如使用刹车系统等等。
Within these advanced automotive camera-based safety systems, enhanced imaging and sensing has opened up a world of possibilities. Camera based vision systems complement long range radar with a long field of view and short range radar or ultrasonic sensor with a wider field of view, allowing the driver to see the road ahead and behind. As a result, another set of data becomes available to confirm the accuracy of the overall picture being drawn. For example, supporting data from radar and cameras provides a significantly higher level of confidence in decision making, such as applying the braking system.
汽车摄像头安全系统内的图像传感器
Image sensors within automotive camera-based safety systems
自1969年电荷耦合器件(CCD)出现以来，其应用一直在增长，扩展到数码相机、望远镜、摄像机、扫描仪，以及其它成像应用领域。然而，CCD在汽车安全应用方面有其固有不足之处。最明显的是：CCD制造需要专门的工厂、设备及工艺，这给以低成本大批量生产汽车摄像头带来了困难。在汽车安全方面，CCD架构也有内在的局限性。标准CCD图像传感器经设计以串行方式读出数据，这意味着必须读出所有的先前像素之后，才能读出后面的像素。这限制了帧速，或每秒读取的图像数目，在快速捕捉图像为关键性功能的汽车安全应用中成为一种重要的缺陷。此外，大多数CCD摄像头不具备表1突显的必需的宽动态范围(WDR)功能。
Since their inception in 1969, charge coupled devices (CCDs) have grown in use. CCDs expanded into digital cameras, telescopes, video cameras, scanners, and other imaging applications. However, CCDs have inherent draw-backs for automotive safety. Most notably, CCD manufacturing requires specialized foundries, equipment, and processes, which prevents them from meeting the low-cost, high-volume production requirements for automotive cameras. The CCD architecture also poses inherent limitations for automotive safety. Standard CCD image sensors are designed for serial readout, meaning that all previous pixels must be read out before the next pixels can be read. Because this limits the frame rate, or number of images read per second, this is a significant disadvantage in automotive safety applications, where capturing images rapidly is a critical function. Additionally, most CCD-based cameras do not provide the necessary wide dynamic range (WDR) as highlighted in Table 1.
表1：汽车摄像头的要求
汽车摄像头要求 描述 传感器要求
高成像灵敏度
或低光照性能 较高的光转换成信号效率，不产生额外的噪声。更高的灵敏度意味着实现具有优良信噪比(SNR)和可用图像所需的光较少。具有高灵敏度的摄像头可以低光照(<0.01 lux)条件下捕捉到良好的图像 需要具备最佳的低光照性能、宽分辨率范围的VGA SOC、WVGA和百万像素传感器。
宽动态范围 在照明条件极端变化下具有高性能。能够在接近前灯、来自其它车辆的眩光、进出隧道、日出和日落等条件下精确捕捉视觉信息，能够缩短曝光响应时间。 具有120dB动态范围的WVGA和百万像素传感器
包括近红外的宽光谱范围灵敏度 近红外(NIR)波长灵敏度可让摄像头捕捉到夜空中出现的波长。也可以通过特殊的前灯来投射近红外光，作为辅助照明，因为人眼是看不到的。 可在白天和夜晚条件下工作
单色和彩色处理
宽光谱范围：400nm到1100nm
近红外灵敏度
高可靠性并符合汽车质量标准 摄像头必须能够耐受汽车使用环境，应当按照ISO/TS16949汽车零部件国际质量管理体系来生产和交货。必须保持高水平的产品质量 生产过程中的严格质量控制
零缺陷
低安全系统总体成本 汽车摄像头统的价格必须对消费者具有吸引力，这是通过控制摄像头自身元件的成本，以及通过摄像头设计来降低总体系统成本的方式来实现 适应性：适合多种应用编程的可配置能力；系统级芯片(SoC)灵活性；降低其它系统元件成本的先进功能
互补金属氧化物半导体(CMOS)图像传感器是替代的解决方案，它可在具有高成本效益的商业代工厂生产，使用与生产其它计算芯片相同的高量产工艺。CMOS图像传感器具有随机访问读取功能，这意味着可对像素进行随机寻址，并可快速读出子帧(sub-frame)，即感兴趣的区域。因此，CMOS图像传感器克服了上述CCD传感器的某些缺点，而且具有几乎无限的子采样和子窗口能力。CMOS摄像头的一个附加优势是能够满足宽动态范围(WDR)和低光照性能要求，进一步推动CMOS图像传感器取代CCD传感器而成为众多成像市场(包括汽车市场)的首选传感器(见图1)。实际上，CMOS汽车传感器市场可望从2009年的不到1000万个增长至2017年的5000万个以上(IMS 2009年汽车市场报告)。
Alternatively, complementary metal oxide semiconductor (CMOS) image sensors are designed for fabrication in cost-effective commercial foundries, using the same high-yielding processes that produce other computing chips. CMOS image sensors have random access readout, meaning that pixels can be randomly addressed and a sub-frame (region of interest) quickly read out. As a result, CMOS image sensors overcome some of the disadvantages of CCDs described above and are nearly unlimited in their ability to sub sample and sub window. As an additional benefit, CMOS cameras are capable of meeting the requirements for WDR and low-light performance, further driving the CMOS image sensors to replace CCDs as the image sensors of choice in numerous imaging markets, including automotive (see automotive application use in Figure 1). In fact, the market for CMOS automotive sensors is forecasted to grow from less than 10 million units in 2009 to more than 50 million units by 2017 (IMS 2009 automotive market report).
在拥有摄像头安全系统车辆内，使用图像传感器来完成场景观察或场景处理。
Within the vehicle’s camera-based safety systems, image sensors are utilized for scene-viewing applications or scene-processing applications.
------------------------------------------
图1 汽车应用中的CMOS图像传感器
场景观察
Scene viewing
通常在视频显示器上输出的场景观察应用包括后视摄像头、多摄像头360度视野、辅助停车、替代后视镜、夜视和后排乘客监视。在这些应用中，关键的图像传感器要求包括出色的低光照性能、WDR和低功耗工作，同时满足汽车质量标准要求，在大批量生产下具有低的单位成本。
Scene-viewing applications which typically output on a video display, include rear view camera, multi-camera 360 degree view, parking assist, wing mirror replacement, night vision and back seat passenger monitoring. In these applications, key image sensor requirements include excellent low-light performance, WDR and low power operation while meeting automotive quality standards and providing low unit costs at high volume.
Aptina imaging公司即是CMOS图像传感器供应商的典范，面向场景观察应用提供全集成式成像解决方案。
例如：专为汽车场景观察系统而设计的VGA系统级芯片(MT9V128和MT9V129)，以及用于下一代视觉摄像头的先进百万像素(MT9M024)解决方案。
Aptina imaging is an example of a CMOS image sensor providor targting scene viewing applicaitons with fully-integrated imaging solutions, such as VGA system on a chip (SOC‘s) designed specifically for automotive scene viewing systems (MT9V128 & MT9V129) and of advanced megapixel (MT9M024) solutions for next generation of viewing cameras.
场景处理
Scene processing
在场景处理应用中，处理单元使用成像器的输出来发送警报或做出有关车辆运行的决策，包括车道偏离警示、交通标志检测、下雨检测、前灯调暗和盲点检测。
Scene-processing applications, where the output of the imager is used by the processing unit to send alerts or make decisions regarding vehicle operation, include lane departure warning, road sign detection, rain detection, headlight dimming and blind spot detection.
使用这些智能化安全系统，处理计算机接收来自传感器的数据，做出决定，并向车辆子系统发送指令，以防止发生事故、减轻事故的严重程度或者保护车主。例如处理计算机可能使用来自图像传感器的数据，确定即将发生的碰撞，并向车辆子系统发出指令以采取刹车措施。
With these intelligent safety system applications, the processing computer receives data from the sensor as input, makes a decision, and sends commands to vehicle subsystems to help prevent an accident, mitigate its severity, or protect the vehicle occupants. For example, a processing computer might use the data from the image sensor to decide that a collision is imminent and send a command to the vehicle subsystems to apply brakes.
微处理器和图像传感器方面的技术进步使得智能化安全系统成为现实。随着技术进步的速度加快，集成更多智能性的图像传感器将成为被动和主动型安全系统的组成部分。
Technological advances in microprocessors and image sensor technologies are making intelligent safety systems a reality. As these advances accelerate, image sensors with integrated intelligence will become part of both passive and active safety systems.
汽车安全性的未来支持
What the future holds in automotive safety
新兴的汽车摄像头安全应用正在以惊人的速度增长，这些应用的配售率一直在上升。地区性标准、交通状况以及驾驶者偏好决定了可用的功能。例如在美国销售的汽车具有后视摄像头、车道偏离警示和360度环绕视觉功能，而欧洲的驾驶者除了360度环绕视觉摄像头外，还能够享有后视摄像头、车道偏离警示、自动远光控制、夜视和交通标志识别功能，每辆车需要多达八个或者更多的摄像头。
Emerging automotive camera-based safety applications are growing at astounding pace and the attach rate for these applications has been on the rise. Regional standards, traffic conditions and driver preferences dictate the features available. For example, cars sold in the U.S. have rear view camera, lane departure warning and or 360 degree surround-view offerings while European drivers see rear view camera, lane departure warning, auto high beam control, night vision and traffic sign recognition features offered in addition to 360 degree surround-view cameras, requiring as many as eight or more cameras per vehicle.
图2 多摄像头360度/环绕视觉系统拓扑
由于每辆车的功能和摄像头数量都在增加，提供集成新功能和新接口的灵活性，对于图像传感器的制造商而言是十分重要的。无论是在摄像头模块内采用协同芯片方式，或者在电子控制单元(ECU)内，最好的办法是使传感器和图像传感器处理ISP功能分开。
As the number of features and the number of cameras per vehicle grow, it is important for image sensor manufacturers to provide the flexibility for the integration of new functions and interfaces. The best approach for achieving this is to separate sensor and Image Sensor Processing ISP function, either within the camera module through a companion chip or in the Electronic Control Unit ECU.
图像传感器制造厂商正在努力提高分辨率，从VGA到百万像素。更高的分辨率可以获得更清晰的街角和边缘。当然也要控制系统的功耗和成本，最大限度地降低摄像头模块的功耗，同时可以获得最佳的图像质量。一种降低成本的方法是使用可用于场景观察和场景处理两种应用的传感器。这就要求根据系统结构、接口和应用状况，在主要针对显示的场景观察和发生在ECU中的场景处理之间复用传感器的输出。Aptina imaging公司开发了解决这些难题的方法，通过提供能够用于摄像头模块和ECU的百万像素传感器以及协同芯片的解决方案，同时实现了更低的功耗和出色的灵活性。
Manufacturers are also working to increase resolution from VGA to megapixel. Higher resolution allows for better definition of corners and edges. Of course, there is also a need to keep power consumption and cost of the system in check. Minimizing power dissipation in the camera module allows for the highest image quality. One way to reduce cost would be to implement sensors that can be utilized in both scene-viewing and scene-processing applications. This requires sensor output to be multiplexed between viewing directed primarily at the display and processing occurring in the ECU, depending on the system architecture, interface and application. An approach to solving these challenges being advanced by Aptina achieves both lower power and excellent flexibility by providing megapixel sensor + companion chip solutions that can be deployed in the camera module and on in the ECU.
结论
Conclusion
汽车业在生产行驶更安全的车辆方面继续承受越来越大的压力。如今，摄像头视觉系统已成为开发更先进安全系统的关键。在这些系统中，CMOS图像传感器不仅能够满足WDR和低光照性能要求，还可提供对于价格敏感的汽车市场至关重要的规模经济效益，有助于该领域实现显著增长。
More and more, the automotive industry is under continued pressure to produce vehicles that are safer to operate. Today, camera-based visibility systems are critical to the development of more advanced safety systems. Within these systems, CMOS image sensors are not only meeting the requirements for WDR and low-light performance, they are also providing the economies of scale critical to the price-sensitive automotive market, contributing to significant growth in the segment.
附件列表
→如果您认为本词条还有待完善，请 编辑词条
词条内容仅供参考，如果您需要解决具体问题
（尤其在法律、医学等领域），建议您咨询相关领域专业人士。
2
收藏到:
中电网官方微博