There are plenty of applications for CMOS image sensors. However, the dynamic range of CMOS image sensor is too narrow to meet the requirement of recording natural scene. Currently, there are a lot of methods to improve the dynamic range; however, most of them have drawbacks, such as deteriorating image quality and recording speed, complicated circuits building blocks, and limited effects on dynamic range, etc. In this proposal, pixel charge compensation method has been introduced to achieve ultra-wide dynamic range CMOS image sensors, creatively. The feature is by inserting another supplementary photodiode in the pixel, which is reverse biasing under weak incident light and is converted automatically to forward biasing when the incident light is strong. This conversion makes it possible to compensate charge to the capacitor of the pixel, to obtain the aim of improving the dynamic range, because the output voltage of the pixels is proportional to the light intensity at weak condition and is proportional to the logarithm of the light intensity under high illumination condition. This strategy has little influence with the process variation. The transient response is very fast at low light level with high signal-to-noise ratio. Under high light level, the output signal doesn't show any saturation phenomenon. Therefore, the dynamic range boosting effect is obvious. The construction of pixels is quite simple, without any complex circuit building blocks and additional timing sequences to judge the incident light level. The system architecture of back-end circuit employs traditional integration readout circuit construction, which is effective, flexible and reliable. The target is achieving ultra-wide CMOS image sensor with close to or above 160dB, at 30 frames per second. The research of this project would solve the issue of the dynamic range of CMOS image sensors perfectly, and it has important scientific significance and application value.
CMOS图像传感器应用广泛,但其动态范围较小,不能满足记录自然场景的需求。目前提高其动态范围的方法存在着牺牲成像速度和质量、电路架构复杂、提升效果有限等缺点。本项目创新地提出利用像素电荷补偿的方法来得到超宽动态范围的CMOS图像传感器,特色是通过像素中加入一个附加的光电二极管,弱光时反偏,强光时自动转换为正偏,从而实现电荷补偿,使得像素输出电压在弱光下与光强成线性,强光下变成与光强对数成正比,达到提高图像传感器动态范围的目的。该方案受工艺影响小,弱光下响应速度快,信噪比高,强光下信号不饱和,动态范围提升明显。像素结构简单,无需采用复杂电路及额外时序来判断入射光的强弱。处理电路及系统方案采用传统积分式结构,技术成熟可靠。目标是所研究的图像传感器在30fps下动态范围将接近或超过160dB。本项目的研究将能够比较完美地解决CMOS图像传感器动态范围的问题,具有重要的科学意义和应用价值。
CMOS图像传感器在消费电子、工业生产、安全监控、医疗成像、科学研究、军事国防等领域都获得了广泛应用,是一类非常重要的集成电路。然而,目前主流传统的CMOS图像传感器动态范围较小,无法记录自然场景接近180dB动态范围的图像信息,因此本项目研发具有超宽动态范围的CMOS图像传感器意义十分重大。.本项目首先提出并研究了基于电荷补偿技术的图像传感器原理;然后利用电荷补偿技术实现了高动态CMOS图像传感器原型器件,对该图像传感器的性能进行了详细的测试,并对非理想因素进行了分析;同时研究了一种基于两次电荷转移的高动态CMOS图像传感器,并研究了高分辨率、高动态范围CMOS图像传感器信号获取和处理电路,设计了图像传感器模拟输出轨对轨放大器,适用于并行列模数转换用的12bit循环型模数转换器,以及适合于串行输出的40MHz采样率、14bit精度的流水线型模数转换器。.本项目提出的像素电荷补偿技术,具有本征的高动态范围,其基本结构是通过像素中加入一个附加的光电二极管,弱光时反偏,强光时自动转换为正偏,从而实现电荷补偿,使得像素输出电压在弱光下与光强成线性,强光下变成与光强对数成正比,达到提高图像传感器动态范围的目的。该技术申请并获得了国家发明专利。.通过这种技术实现的CMOS图像传感器像素在弱光下呈现线性响应,在强光下为对数响应,可响应光功率密度在1.99×10^-9W/cm^2至0.448W/cm^2之间,等效动态范围达到167dB,远远高于目前主流CMOS图像传感器约80dB的动态范围。同时该图像传感器具有接近2V的输出摆幅。设计的14bit流水线模数转换器采用了每级2.5bit精度的冗余架构,前期测量结果显示有效精度超过了10bit。.本项目所研究的图像传感器结构简单、处理电路与信号处理方案成熟可靠,将能够比较完美地解决传统CMOS图像传感器动态范围较低的缺点,具有重要的科学意义和应用价值。
{{i.achievement_title}}
数据更新时间:2023-05-31
路基土水分传感器室内标定方法与影响因素分析
涡度相关技术及其在陆地生态系统通量研究中的应用
环境类邻避设施对北京市住宅价格影响研究--以大型垃圾处理设施为例
内点最大化与冗余点控制的小型无人机遥感图像配准
滴状流条件下非饱和交叉裂隙分流机制研究
基于深亚微米工艺大动态范围、千万像素CMOS图像传感器关键技术研究
20M像素高速高动态范围CMOS图像传感器关键技术研究
基于像素单元的CMOS图像传感器辐射损伤机理研究
基于单光子计数CMOS有源像素的量子图像传感器研究