WO2020082392A1 - 图像传感器及成像方法 - Google Patents

图像传感器及成像方法 Download PDF

Info

Publication number
WO2020082392A1
WO2020082392A1 PCT/CN2018/112266 CN2018112266W WO2020082392A1 WO 2020082392 A1 WO2020082392 A1 WO 2020082392A1 CN 2018112266 W CN2018112266 W CN 2018112266W WO 2020082392 A1 WO2020082392 A1 WO 2020082392A1
Authority
WO
WIPO (PCT)
Prior art keywords
cavity
pixel
image sensor
band
sensor array
Prior art date
Application number
PCT/CN2018/112266
Other languages
English (en)
French (fr)
Inventor
王星泽
赖嘉炜
舒远
Original Assignee
合刃科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 合刃科技（深圳）有限公司 filed Critical 合刃科技（深圳）有限公司
Priority to PCT/CN2018/112266 priority Critical patent/WO2020082392A1/zh
Priority to CN201880068674.1A priority patent/CN111356649B/zh
Publication of WO2020082392A1 publication Critical patent/WO2020082392A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B81—MICROSTRUCTURAL TECHNOLOGY
    • B81B—MICROSTRUCTURAL DEVICES OR SYSTEMS, e.g. MICROMECHANICAL DEVICES
    • B81B7/00—Microstructural systems; Auxiliary parts of microstructural devices or systems
    • B81B7/02—Microstructural systems; Auxiliary parts of microstructural devices or systems containing distinct electrical or optical devices of particular relevance for their function, e.g. microelectro-mechanical systems [MEMS]
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
    • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
    • H01L27/14—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components sensitive to infrared radiation, light, electromagnetic radiation of shorter wavelength or corpuscular radiation and specially adapted either for the conversion of the energy of such radiation into electrical energy or for the control of electrical energy by such radiation
    • H01L27/144—Devices controlled by radiation
    • H01L27/146—Imager structures
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
    • H04N1/04—Scanning arrangements, i.e. arrangements for the displacement of active reading or reproducing elements relative to the original or reproducing medium, or vice versa
    • H04N1/203—Simultaneous scanning of two or more separate pictures, e.g. two sides of the same sheet

Definitions

  • the invention relates to the field of image sensing technology, in particular to an image sensor and imaging method.
  • multi-spectral sensors can use light from different wavelengths to detect objects to obtain more information.
  • filter structures such as a filter wheel based on a dye filter, a liquid crystal tunable filter (Liquid Crystal Tunable Filter, or LCTF for short) based on a Lyot filter group ) And Acousto-optical Tunable Filter (AOTF) based on the elastic-optic effect.
  • AOTF Acousto-optical Tunable Filter
  • the three-color dye filter in front of the original pixel array is replaced with a reflective filter structure formed by a Fabry-Pérot (FP) cavity.
  • FP Fabry-Pérot
  • the inventors found through research that in the prior art, due to the silicon-based complementary metal oxide semiconductor (Complementary Metal-Oxide-Semiconductor, CMOS for short) image sensor, the quantum efficiency of the image sensor fluctuates significantly at different wavelengths. To ensure that the image sensor has a close response to all bands to avoid large area saturation of pixels or below noise level, it is necessary to optimize the transmittance of the filter structure in the prior art to balance the energy of each band.
  • CMOS complementary Metal-Oxide-Semiconductor
  • an image sensor including:
  • a pixel sensor array and a FP cavity reflective filter structure located at the front of the pixel sensor array; wherein the FP cavity reflective filter structure includes a bottom reflector, an FP cavity, and a top reflector; the structure of the image sensor The pixel sensor array, the bottom mirror, the FP cavity, and the top mirror are sequentially arranged from bottom to top; wherein, the structural properties and parameters of the FP cavity reflective filter structure can be adjusted.
  • the structural properties and parameters that can be adjusted include the cavity length of the FP cavity, the cavity medium of the FP cavity, the thickness of the bottom mirror and the top mirror, the bottom layer The material of the reflector and the top reflector.
  • the F-P cavity reflective filter structure corresponding to the pixel sensor array and its front end is arranged in a mosaic, wherein each pixel has a unique combination of wavelength band and exposure intensity.
  • a balanced binary tree model is used to perform segmentation processing on the band and the exposure intensity of each pixel.
  • the material of the bottom mirror and the top mirror is aluminum or silver.
  • the pixel sensor array is a CMOS sensor array.
  • the first step is to obtain multiple low-spatial-resolution images with multiple bands and multiple exposure intensities using the single-frame image acquired by the image sensor;
  • the second step is to perform high dynamic range synthesis processing on multiple exposure images of each band, and use the camera response function to normalize the pixel values under different exposure intensities; repeat the same operation for each band to obtain each High dynamic range pixel values for each band;
  • the pixels in the neighborhood are used to perform interpolation processing to recover, and a high dynamic range image after resolution recovery is generated.
  • the neighborhood of the pixel to be interpolated is expanded, and the weight of different pixels in the neighborhood in the fitting process is determined by calculating the gradient change of the pixel value.
  • the FP cavity reflective filter can cover a wide range of bands in real time, densely and continuously, and has the characteristics of narrow band and high transmittance.
  • the FP cavity with different media and thickness and the mirrors of different materials and thickness can solve each band Under-exposure or over-exposure problems, improve the utilization of pixels and the dynamic range of captured images.
  • FIG. 1 is a schematic structural diagram of an image sensor in the present invention
  • FIG. 2 is a schematic diagram of a filter design for energy balance in the visible light band in the present invention
  • FIG. 3 is a schematic diagram of dividing the band and the exposure intensity by using the balanced binary tree model in the present invention
  • FIG. 4 is a schematic diagram of the arrangement of the pixel point sensor array and the F-P cavity reflective filter structure corresponding to the front end of the invention
  • FIG. 5 is a flowchart of a high dynamic range image imaging method of the present invention.
  • the embodiment of the present invention discloses an image sensor, and particularly relates to a chip integrated real-time hyperspectral high dynamic range image sensor.
  • the semiconductor thin film technology is used to directly deposit the F-P cavity thin film array on the pixel sensor array of the image sensor to replace the traditional RGB dye filter.
  • FIG. 1 it is a schematic structural diagram of an image sensor in the present invention.
  • a bottom mirror 2 is provided above the pixel sensor array 3, and an FP cavity 4, FP is provided above the bottom mirror 2
  • a top mirror 1 is provided above the cavity 4, that is, the structure of the image sensor is a pixel sensor array 3, a bottom mirror 2, a FP cavity 4, and a top mirror 1 in sequence from bottom to top; wherein, the bottom layer The reflecting mirror 2, the FP cavity 4, and the top reflecting mirror 1 form an FP cavity reflecting filter structure; wherein, when the bottom reflecting mirror 2, the top reflecting mirror 1 are made of metal materials, Then, the bottom reflecting mirror 2 and the top reflecting mirror 1 form a metal reflecting layer of the FP cavity reflective filter structure.
  • the image sensor is a CMOS image sensor.
  • the wavelength and intensity of transmitted light can be controlled, so that the image sensor can obtain a uniform response in each band.
  • the structural parameters and properties that can be changed include but are not limited to the cavity length of the FP cavity 4, the cavity medium of the FP cavity 4 (different cavity media have different refractive indexes), the bottom mirror 2 and the top layer The thickness of the mirror 1, the materials of the bottom mirror 2 and the top mirror 1, etc.
  • the cavity medium deposition thickness of the FP cavity 4 is changed by changing the cavity medium deposition time of the FP cavity 4 in front of each pixel in the pixel sensor array 3, so that the cavity length of the FP cavity 4 can meet the target wavelength Constructive interference conditions, so as to achieve the selection of specific wavelengths of light.
  • changing the optical length of the FP cavity 4 includes, but is not limited to, changing the cavity length of the FP cavity 4, replacing the cavity medium with different transmittances, modulating the cavity medium with a variable refractive index, etc.
  • the choice of wavelength includes increasing the cavity length of the F-P cavity 4 or using a cavity medium with a higher refractive index can extend the filter wavelength of the F-P cavity to the near-infrared region, further enhancing the hyperspectral acquisition capability of the image sensor.
  • reducing the thickness of the bottom mirror 2 or the top mirror 1 can reduce the reflectivity of the mirror.
  • the bottom reflector 2 and the top reflector 1 are made of metal materials, the bottom reflector 2 and the top reflector 1 form a metal reflection layer of the FP cavity reflective filter structure; the thickness of the metal reflection layer is reduced
  • the reflectivity of the metal reflective layer can be reduced, and the transmittance of the FP cavity 4 and the half-wave width of the transmission band can be increased.
  • the quantum efficiency of the CMOS image sensor in the selected band can be improved, so that the normalized response of each band of the CMOS image sensor is at the same level.
  • the use of different metal materials to make the bottom mirror 2 and the top mirror 1 can change the transmittance of the FP cavity reflective filter structure to different bands of light; when the bottom mirror 2, the top mirror 1 is made of metal Made of materials, the bottom mirror 2 and the top mirror 1 form a metal reflective layer of the FP cavity reflective filter structure.
  • different metals can be used to prepare the metal reflective layer of the F-P cavity reflective filter structure.
  • aluminum (Al) can be used as the metal reflective layer material of the FP cavity reflective filter structure. Since aluminum (Al) has a low reflectivity in the near infrared band, it can compensate for the CMOS image sensor in the near infrared The lower band quantum efficiency achieves energy balance.
  • silver (Ag) can be used as the material of the metal reflective layer of the FP cavity reflective filter structure, because the light reflectivity of silver (Ag) in the range below 450 nm is low, which makes the FP cavity in the short wavelength band A higher transmittance can be obtained, and the energy balance is further improved, so that the normalized response M n of the image sensor in each wave band is approximately equal.
  • FIG. 2 is a schematic diagram of a filter design for energy balance in the visible light band of the present invention.
  • multiple F-P cavity film arrays with different transmission wavelengths and different transmission intensities can be obtained on the image sensor.
  • Multi-dimensional image information including spatial position, wavelength, brightness, etc. can be obtained after single-frame shooting with the image sensor.
  • the additional added pixel information generated by the FP cavity reflective filter structure with multiple transmission rates in the same waveband You can obtain the image information of the subject under different exposure intensities.
  • the low dynamic range (LDR) images obtained by the subject under a single brightness are synthesized to obtain images with high dynamic range (HDR), which can enhance the sensor's large brightness span.
  • HDR high dynamic range
  • the ability to capture light and dark details in the scene especially the ability to distinguish reflective water bodies, roads, forest shadows and other areas. Since the pixel sensor array is divided by multiple different wavelength bands and exposure intensity, in order to achieve the ideal image quality, each sub-image must be fused to restore the spatial resolution of the image.
  • FIG. 3 it is a schematic diagram of using the balanced binary tree model to separately divide the band and the exposure intensity in the present invention. By dividing the band and the exposure intensity separately to ensure that each band or the exposure intensity is not discriminated; wherein, FIG. 3 A / B / ... in the table shows the difference in exposure caused by different transmittances, and 1/2/3/4 / ... stands for different bands.
  • the pixel sensor array of the image sensor in the present invention is designed according to a mosaic arrangement, and the FP cavity reflective filter structure at the front of the pixel sensor array also corresponds to the pixel sensor array arranged according to a mosaic.
  • a / B / ... represents the difference in exposure caused by different transmittances
  • 1/2/3/4 / ... represents different wavebands
  • each pixel is distributed as uniformly as possible In the entire image sensor area, to avoid the deterioration of a certain imaging area due to sparse sampling; wherein, each pixel has a unique combination of band and exposure intensity.
  • the length of the balanced binary tree of the sub-band and the sub-exposure can be separately expanded and contracted to change the number of band channels. For example, when the image sensor is applied to a scene with a large brightness range, the number of exposure condition channels is appropriately increased. Or, when the image sensor is used in the detection industry, in order to improve the frequency domain resolution, or to expand the detection range to a wider band range, the number of exposure channels can be reduced and assigned to a new band channel.
  • the embodiment of the present invention discloses an imaging method, and particularly relates to an imaging method for high dynamic range images.
  • FIG. 5 it is a flow chart of a method for realizing high dynamic range image imaging using the image sensor in the present invention:
  • the first step is to use the single-frame image obtained by the image sensor to obtain several low-spatial resolution images in multiple bands and multiple exposure intensities;
  • the second step is to perform high dynamic range (HDR) synthesis processing on multiple exposure images of each band, and use the camera response function to normalize the pixel values under different exposure intensities to improve the dynamic range of the image; Repeat the same operation for each band to get the high dynamic range (HDR) pixel value for each band;
  • HDR high dynamic range
  • the pixel interpolation in the neighborhood is used to restore the resolution to generate a high dynamic range image after the resolution is restored.
  • the pixels in the neighborhood all contain different exposure intensity and band information; because the pixels of each band are sparsely distributed on the image sensor, in order to effectively perform the resolution restoration process, it is necessary Expand the neighborhood of the pixel to be interpolated, and calculate the gradient change of the pixel value to determine the weight of different pixels in the neighborhood in the fitting process.
  • the FP cavity reflective filter structure is used to replace the three-color RGB dye filter in the traditional image sensor for light wavelength selection, and the transmission wavelength is selected by changing the optical length of the FP cavity; by changing the FP cavity reflective filter
  • the material and thickness of the bottom mirror and the top mirror (metal reflective layer) in the optical structure, adjust the transmittance of the FP cavity to achieve energy balance, and set multiple transmittance gears for each band to obtain multiple Low-dynamic-range LDR images require only a single frame to capture high-spectrum high-dynamic-range HDR images.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Power Engineering (AREA)
  • Multimedia (AREA)
  • Electromagnetism (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Solid State Image Pick-Up Elements (AREA)
  • Studio Devices (AREA)
  • Transforming Light Signals Into Electric Signals (AREA)

Abstract

本发明实施例公开了一种图像传感器，包括：像素点传感器阵列、位于像素点传感器阵列前端的F-P腔反射型滤光结构；其中，所述F-P腔反射型滤光结构包括底层反射镜、F-P腔体、顶层反射镜；所述图像传感器的结构从下至上依次为像素点传感器阵列、底层反射镜、F-P腔体、顶层反射镜；其中，F-P腔反射型滤光结构的结构属性及参数可以调节。此外，本发明实施例还公开了一种高动态范围图像成像方法。采用本发明中的图像传感器及成像方法，解决了在成像过程中各个波段欠曝或过曝的问题，提高了像素利用率和拍摄图像的动态范围。

Description

图像传感器及成像方法 技术领域
本发明涉及图像传感技术领域，特别涉及一种图像传感器及成像方法。
背景技术
现有技术中，基于对光电器件的丰富研究成果，工程师已经制造出各种各样适用于工业、农业、矿业以及监控领域的图像传感器及相应的信号处理系统。而区别于传统的黑白或RGB传感器，多光谱传感器可利用不同波段的光线来探测物体以获得更多信息。通过在常规传感器前加装不同的滤波结构，可以获得多个不同波段的窄带光信号，例如基于染料滤色镜的滤波轮、基于Lyot滤镜组的液晶可调滤镜(Liquid Crystal Tunable Filter，简称LCTF)和基于弹光效应的声光可调滤镜(Acousto-optical Tunable Filter，简称AOTF)。然而，发明人经研究发现，上述现有技术中的多光谱传感器均存在体积庞大、整体捕获时间长、滤光带宽较宽的问题。
目前，在传统的Bayer式RGB传感器基础上，将原来像素阵列前的三色染料型滤波片更换为法布里-佩罗(Fabry-Pérot，简称F-P)腔构成的反射型滤光结构，通过改变反射腔的长度可以获得不同波段的窄带滤波效果。但是，发明人经研究发现，在现有技术中，由于硅基互补金属氧化物半导体(Complementary Metal-Oxide-Semiconductor，简称CMOS)图像传感器的量子效率在不同波长条件下存在明显起伏的情况，为了保证图像传感器对所有的波段具有接近的响应以避免像素大面积饱和或低于噪声水平，有必要对现有技术中滤波结构的透过率进行优化而使得各波段的能量达到平衡。
发明内容
基于此，为解决现有技术中的技术问题，特提出了一种图像传感器，包括：
像素点传感器阵列、位于像素点传感器阵列前端的F-P腔反射型滤光结构；其中，所述F-P腔反射型滤光结构包括底层反射镜、F-P腔体、顶层反射镜；所述图像传感器的结构从下至上依次设置为像素点传感器阵列、底层反射镜、F-P 腔体、顶层反射镜；其中，所述F-P腔反射型滤光结构的结构属性及参数可以调节。
在一个实施例中，可以调节的所述结构属性及参数包括所述F-P腔体的腔长、F-P腔体的腔体介质、所述底层反射镜及所述顶层反射镜的厚度、所述底层反射镜及所述顶层反射镜的材质。
在一个实施例中，所述像素点传感器阵列及其前端对应的所述F-P腔反射型滤光结构按照马赛克排布，其中，每一个像素点具有唯一的波段和曝光强度组合。
在一个实施例中，利用平衡二叉树模型对所述每一个像素点的所述波段和所述曝光强度进行分割处理。
在一个实施例中，所述底层反射镜及所述顶层反射镜的材质为铝或者银。
在一个实施例中，所述像素点传感器阵列为CMOS传感器阵列。
此外，为解决现有技术中的技术问题，特提出了一种高动态范围图像的成像方法：
第一步，利用图像传感器获取的单帧图像获得多个波段、多个曝光强度下的多个低空间分辨率图像；
第二步，对每个波段的多个曝光图像进行高动态范围合成处理，并利用相机响应函数将不同曝光强度下的像素值进行归一化处理；对每个波段重复相同的操作，得到每个波段的高动态范围像素值；
第三步，对于因为像素饱和或因为低于噪声水平而丢失的像素，利用邻域内的像素进行插值处理来恢复，生成分辨率恢复后的高动态范围图像。
在一个实施例中，扩大待插值像素点的邻域，并且通过计算像素值的梯度变化以确定邻域内不同像素点在拟合处理中的权重。
实施本发明实施例，将具有如下有益效果：
F-P腔反射型滤光片可以实时、密集、连续覆盖大范围的波段，并且具有窄带高透过率的特性，不同介质、厚度的F-P腔体配合不同材质、厚度的反射镜，可以解决各个波段中欠曝或过曝问题，提高像素的利用率和拍摄图像的动态范围。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
其中：
图1为本发明中的图像传感器结构示意图；
图2为本发明中可见光波段内能量平衡的滤光片设计示意图；
图3为本发明中利用平衡二叉树模型分别对波段和曝光强度进行分割的示意图；
图4为本发明中像素点传感器阵列及其前端对应的F-P腔反射型滤光结构的排布示意图；
图5为本发明中高动态范围图像成像方法的流程图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明的实施例公开了一种图像传感器，特别涉及了一种芯片集成式的实时高光谱高动态范围图像传感器。在本发明的实施例中，利用半导体薄膜技术，直接在图像传感器的像素点传感器阵列上沉积F-P腔薄膜阵列，以替代传统的RGB染料滤光片。如图1所示为本发明中的图像传感器结构示意图，在所述图像传感器中，像素点传感器阵列3的上方设置有底层反射镜2，底层反射镜2的上方设置有F-P腔体4，F-P腔体4的上方设置有顶层反射镜1，即所述图像传感器的结构从下至上依次为像素点传感器阵列3、底层反射镜2、F-P腔体4、顶层反射镜1；其中，所述底层反射镜2、所述F-P腔体4、所述顶层反射镜1组成了F-P腔反射型滤光结构；其中，当所述底层反射镜2、所述顶层反射镜1 由金属材料制作而成，则所述底层反射镜2、所述顶层反射镜1组成了F-P腔反射型滤光结构的金属反射层。
在一种实施例中，图像传感器为CMOS图像传感器。
如图1所示，通过改变像素点传感器阵列3前端的F-P腔反射型滤光结构的结构属性及参数可以控制透射光的波长及强度，使图像传感器在各个波段下获得均匀的响应。
其中，可以改变的所述结构参数及属性包括但不仅限于F-P腔体4的腔长、F-P腔体4的腔体介质(不同的腔体介质具有不同的折射率)、底层反射镜2及顶层反射镜1的厚度、底层反射镜2及顶层反射镜1的材质等。
通过改变像素点传感器阵列3中各个像素点前端F-P腔体4的腔体介质沉积时间来改变所述F-P腔体4的腔体介质沉积厚度，使得F-P腔体4的腔长能够满足目标波长的相长干涉条件，从而实现对特定波长的光进行选择。
其中，改变F-P腔体4的光学长度，包括但不限于改变F-P腔体4的腔长、更换不同透射率的腔体介质、对折射率可变的腔体介质进行调制等，从而实现对透射波长的选择。其中，增加F-P腔体4的腔长，或利用具有更高折射率的腔体介质，可以将F-P腔的滤波波长拓展至近红外区域，进一步增强图像传感器的高光谱采集能力。
其中，减少底层反射镜2或者顶层反射镜1的厚度可以降低反射镜的反射率。当利用金属材质制作底层反射镜2及顶层反射镜1时，则所述底层反射镜2、所述顶层反射镜1组成了F-P腔反射型滤光结构的金属反射层；减少金属反射层的厚度可以降低金属反射层的反射率，增加F-P腔体4的透射率和透过波段的半波宽度。利用这一原理可以提高CMOS图像传感器在选定波段的量子效率，使CMOS图像传感器各波段的归一化响应处于同一水平上。
其中，利用不同的金属材料制作底层反射镜2及顶层反射镜1可以改变F-P腔反射型滤光结构对不同波段光线的透射率；当所述底层反射镜2、所述顶层反射镜1由金属材料制作而成，则所述底层反射镜2、所述顶层反射镜1组成了F-P腔反射型滤光结构的金属反射层。针对不同的波段，可以利用不同的金属制备所述F-P腔反射型滤光结构的金属反射层。
在一个实施例中，可以采用铝(Al)作为F-P腔反射型滤光结构的金属反射层材质，由于铝(Al)在近红外波段具有较低的反射率，可以补偿CMOS图像传感器在近红外波段较低的量子效率以达到能量平衡。
在一个实施例中，可以采用银(Ag)作为F-P腔反射型滤光结构的金属反射层材质，由于银(Ag)在450nm以下范围内的光线反射率较低，从而使得F-P腔在短波段内能够获得较高的透射率，并进一步完善能量平衡，使图像传感器在各个波段的归一化响应M n大致相等。
所述归一化响应M n的计算公式如下：
Figure PCTCN2018112266-appb-000001
其中λ为波长，n为波段通道数，I(λ)为不同照明场合下的能量分布，R(λ)为裸传感器在各个波段下的响应，T n为各波段通道滤光片的透射率函数。此外，对于透射率函数T n，利用F-P腔体的分布函数代替染料滤光片常用的高斯分布函数来进行优化设计。如图2所示为本发明中可见光波段内能量平衡的滤光片设计示意图。
通过调整所述F-P腔反射型滤光结构的结构参数及属性，可以在图像传感器上获得多个不同透射波长、不同的透射强度的F-P腔薄膜阵列。利用所述图像传感器进行单帧拍摄后便可以获取包括空间位置、波长、亮度等的多维图像信息。
当所述图像传感器应用在遥感拍摄场景下时，除了能够实时地获取景物的高光谱信息外，利用额外加入的同一波段内具有多档透射率的F-P腔反射型滤光结构所产生的像素信息可以获取被拍摄物在不同曝光强度下的图像信息。将被拍摄物在单个亮度下获取的低动态范围(Low Dynamic Range，简称LDR)图像进行合成，获得具有高动态范围(High Dynamic Range，简称HDR)的图像，可以增强传感器在较大亮度跨度的场景下捕获明暗细节的能力，尤其是对反光的水体、道路、森林阴影等区域的分辨能力。由于像素点传感器阵列被多个不同波段和曝光强度分割，为了达到理想的图像质量，必须对各个子图像进行融 合，进而恢复图像的空间分辨率。
如图3所示为本发明中利用平衡二叉树模型分别对波段和曝光强度进行分割的示意图，通过分别对波段和曝光强度进行分割以保证每个波段或曝光强度都不被歧视；其中，图3中的A/B/…表示不同的透射率造成的曝光差异，1/2/3/4/…代表不同的波段。
如图4所示，本发明中图像传感器的像素点传感器阵列按照马赛克排布设计，所述像素点传感器阵列前端的F-P腔反射型滤光结构也对应于像素点传感器阵列按照马赛克进行排布，其中A/B/…表示不同的透射率造成的曝光差异，1/2/3/4/…代表不同的波段，为了保证每种像素点的空间分布均匀性，每种像素点尽可能均匀分布在整个图像传感器区域，以避免因为稀疏采样导致某个成像区域的劣化；其中，每一个像素点具有唯一的波段和曝光强度组合。
在一个实施例中，针对不同的应用场合，可以分别对分波段和分曝光的平衡二叉树的长度进行伸缩以改变波段通道数。例如，当将该图像传感器应用于亮度范围大的场景时，则适当增加曝光条件通道数。或者，当将该图像传感器应用于检测行业中，为了提高频域分辨率，或为了将探测范围拓展至更宽的波段范围，可以减少曝光通道数并将其分配到新的波段通道上。
本发明的实施例公开了一种成像方法，特别涉及了一种高动态范围图像的成像方法。如图5所示为本发明中利用所述图像传感器实现高动态范围图像成像方法的流程图：
第一步，利用图像传感器获取的单帧图像获得多个波段、多个曝光强度下的若干低空间分辨率图像；
第二步，对每一个波段的多个曝光图像进行高动态范围(HDR)合成处理，并利用相机响应函数将不同曝光强度下的像素值进行归一化处理，以提高图像的动态范围；对每个波段重复相同的操作，得到每个波段的高动态范围(HDR)像素值；
第三步，对于因为像素饱和或因为低于噪声水平而丢失的像素，利用邻域内的像素插值来恢复分辨率，生成分辨率恢复后的高动态范围图像。
其中，对于待插值的像素点，其邻域内的像素均包含了不同的曝光强度及波段信息；由于每个波段的像素点稀疏地分布在图像传感器上，为了有效地进 行分辨率恢复处理，需要扩大待插值像素点的邻域，并且通过计算像素值的梯度变化以确定邻域内不同像素点在拟合处理中的权重。
实施本发明实施例，将具有如下有益效果：
采用F-P腔反射型滤光结构代替传统图像传感器中的三色RGB染料滤光片对光进行波长选择，通过改变F-P腔体的光学长度来实现对透射波长的选择；通过改变F-P腔反射型滤光结构中底层反射镜及顶层反射镜(金属反射层)的材质和厚度，对F-P腔的透射率进行调整以达到能量平衡，同时对各个波段均设置多个透射率档位以同时获得多个低动态范围LDR图像，仅仅需要单帧拍摄便可以实现高光谱高动态范围HDR图像的拍摄。
以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不会使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (8)

  1. 一种图像传感器，其特征在于，包括：
    像素点传感器阵列、位于所述像素点传感器阵列前端的F-P腔反射型滤光结构；其中，所述F-P腔反射型滤光结构包括底层反射镜、F-P腔体、顶层反射镜；所述图像传感器的结构从下至上依次设置为像素点传感器阵列、底层反射镜、F-P腔体、顶层反射镜；其中，所述F-P腔反射型滤光结构的结构属性及参数可以调节。
  2. 根据权利要求1所述的图像传感器，其特征在于，
    可以调节的所述结构属性及参数包括所述F-P腔体的腔长、所述F-P腔体的腔体介质、所述底层反射镜及所述顶层反射镜的厚度、所述底层反射镜及所述顶层反射镜的材质。
  3. 根据权利要求2所述的图像传感器，其特征在于，
    所述像素点传感器阵列及其前端对应的所述F-P腔反射型滤光结构按照马赛克排布，其中，每一个像素点具有唯一的波段和曝光强度组合。
  4. 根据权利要求3所述的图像传感器，其特征在于，
    利用平衡二叉树模型对所述每一个像素点的所述波段和所述曝光强度进行分割处理。
  5. 根据权利要求2或3或4所述的图像传感器，其特征在于，
    所述底层反射镜及所述顶层反射镜的材质为铝或者银。
  6. 根据权利要求2或3或4所述的图像传感器，其特征在于，
    所述像素点传感器阵列为CMOS传感器阵列。
  7. 一种成像方法，其特征在于，包括：
    第一步，利用图像传感器获取的单帧图像获得多个波段、多个曝光强度下的多个低空间分辨率图像；
    第二步，对每个波段的多个曝光图像进行高动态范围合成处理，并利用相机响应函数将不同曝光强度下的像素值进行归一化处理；对每个波段重复相同的操作，得到每个波段的高动态范围像素值；
    第三步，对于因为像素饱和或因为低于噪声水平而丢失的像素，利用邻域内的像素进行插值处理来恢复，生成分辨率恢复后的高动态范围图像。
  8. 根据权利要求7所述的装置，其特征在于，
    扩大待插值像素点的邻域，并且通过计算像素值的梯度变化以确定邻域内不同像素点在拟合处理中的权重。
PCT/CN2018/112266 2018-10-27 2018-10-27 图像传感器及成像方法 WO2020082392A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2018/112266 WO2020082392A1 (zh) 2018-10-27 2018-10-27 图像传感器及成像方法
CN201880068674.1A CN111356649B (zh) 2018-10-27 2018-10-27 图像传感器及成像方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/112266 WO2020082392A1 (zh) 2018-10-27 2018-10-27 图像传感器及成像方法

Publications (1)

Publication Number Publication Date
WO2020082392A1 true WO2020082392A1 (zh) 2020-04-30

Family

ID=70332102

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/112266 WO2020082392A1 (zh) 2018-10-27 2018-10-27 图像传感器及成像方法

Country Status (2)

Country Link
CN (1) CN111356649B (zh)
WO (1) WO2020082392A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113093322B (zh) * 2021-03-30 2023-03-28 联合微电子中心有限责任公司 Cmos图像传感器、干涉型滤光片及其制备方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040234873A1 (en) * 2002-05-13 2004-11-25 Kartik Venkataraman Color filter imaging array and method of formation
CN102037717A (zh) * 2008-05-20 2011-04-27 派力肯成像公司 使用具有异构成像器的单片相机阵列的图像拍摄和图像处理
CN102437166A (zh) * 2011-10-09 2012-05-02 中国科学院苏州纳米技术与纳米仿生研究所 一种非制冷红外探测系统像素阵列的制作方法
CN104535007A (zh) * 2014-12-15 2015-04-22 哈尔滨工程大学 一种基于腔长可调f-p白光干涉解调装置的分布式光纤应变测量系统
CN105450937A (zh) * 2014-08-29 2016-03-30 展讯通信（上海）有限公司 图像处理方法和图像处理器
CN106672893A (zh) * 2016-12-29 2017-05-17 深圳先进技术研究院 可调谐滤波器以及可调谐滤波器阵列

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008128372A1 (en) * 2007-04-20 2008-10-30 Eth Zurich Transmission interferometric adsorption sensor
EP2522968B1 (en) * 2009-11-30 2021-04-21 IMEC vzw Integrated circuit for spectral imaging system
CN106768324A (zh) * 2016-11-17 2017-05-31 天津津航技术物理研究所 一种光谱成像微型传感器

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040234873A1 (en) * 2002-05-13 2004-11-25 Kartik Venkataraman Color filter imaging array and method of formation
CN102037717A (zh) * 2008-05-20 2011-04-27 派力肯成像公司 使用具有异构成像器的单片相机阵列的图像拍摄和图像处理
CN102437166A (zh) * 2011-10-09 2012-05-02 中国科学院苏州纳米技术与纳米仿生研究所 一种非制冷红外探测系统像素阵列的制作方法
CN105450937A (zh) * 2014-08-29 2016-03-30 展讯通信（上海）有限公司 图像处理方法和图像处理器
CN104535007A (zh) * 2014-12-15 2015-04-22 哈尔滨工程大学 一种基于腔长可调f-p白光干涉解调装置的分布式光纤应变测量系统
CN106672893A (zh) * 2016-12-29 2017-05-17 深圳先进技术研究院 可调谐滤波器以及可调谐滤波器阵列

Also Published As

Publication number Publication date
CN111356649B (zh) 2023-10-13
CN111356649A (zh) 2020-06-30

Similar Documents

Publication Publication Date Title
CN108419061B (zh) 基于多光谱的图像融合设备、方法及图像传感器
US9823128B2 (en) Multispectral imaging based on computational imaging and a narrow-band absorptive filter array
EP3031202B1 (en) Dynamic color shading correction
US8934034B2 (en) Generalized assorted pixel camera systems and methods
EP2420051B1 (en) Producing full-color image with reduced motion blur
US20060013479A1 (en) Restoration of color components in an image model
US20090190853A1 (en) Image noise reduction apparatus and method, recorded medium recorded the program performing it
CN105891928B (zh) 一种日夜兼用的摄像滤波器
JP4328424B2 (ja) 画像変換方法
CN109716758A (zh) 使用对能见度状态的改变具有鲁棒性的复合滤波法的监控相机和采用其的视频监控系统
CN113884184B (zh) 使用光学滤波器子阵列的光谱传感器系统
JP2001099710A5 (zh)
CN113676628A (zh) 多光谱传感器、成像装置和图像处理方法
JP2008244833A (ja) 撮像装置、色収差補正方法およびプログラム
CN111999788A (zh) 宽光谱编码全彩色滤光片阵列
WO2020082392A1 (zh) 图像传感器及成像方法
FR2899696A1 (fr) Procede de traitement d'un phenomene d'eclairement relatif sur une image numerique et systeme de traitement associe
US11696043B2 (en) White balance compensation using a spectral sensor system
CN110999283B (zh) 成像装置、成像方法以及存储介质
Monno et al. N-to-sRGB mapping for single-sensor multispectral imaging
WO2022020989A1 (zh) 一种滤光阵列、移动终端以及设备
CN115000107A (zh) 多光谱成像芯片、多光谱成像组件、制备方法及移动终端
Asiq et al. Efficient colour filter array demosaicking with prior error reduction
CN205622731U (zh) 一种图像传感器
WO2022194040A1 (zh) 基于光谱芯片的图像传感方法、光谱恢复方法、装置和电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18938063

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08.09.2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18938063

Country of ref document: EP

Kind code of ref document: A1