WO2019144678A1 - 成像元件、成像设备和图像信息处理方法 - Google Patents

成像元件、成像设备和图像信息处理方法 Download PDF

Info

Publication number
WO2019144678A1
WO2019144678A1 PCT/CN2018/115225 CN2018115225W WO2019144678A1 WO 2019144678 A1 WO2019144678 A1 WO 2019144678A1 CN 2018115225 W CN2018115225 W CN 2018115225W WO 2019144678 A1 WO2019144678 A1 WO 2019144678A1
Authority
WO
WIPO (PCT)
Prior art keywords
current
pixel
imaging element
scaling
pixel processing
Prior art date
Application number
PCT/CN2018/115225
Other languages
English (en)
French (fr)
Inventor
冯杰
Original Assignee
北京灵汐科技有限公司
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 北京灵汐科技有限公司 filed Critical 北京灵汐科技有限公司
Priority to JP2020522333A priority Critical patent/JP6918390B2/ja
Publication of WO2019144678A1 publication Critical patent/WO2019144678A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/70—SSIS architectures; Circuits associated therewith
    • H04N25/703—SSIS architectures incorporating pixels for producing signals other than image signals
    • H04N25/708—Pixels for edge detection
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/70—SSIS architectures; Circuits associated therewith
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/40—Extracting pixel data from image sensors by controlling scanning circuits, e.g. by modifying the number of pixels sampled or to be sampled
    • H04N25/46—Extracting pixel data from image sensors by controlling scanning circuits, e.g. by modifying the number of pixels sampled or to be sampled by combining or binning pixels
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/70—SSIS architectures; Circuits associated therewith
    • H04N25/76—Addressed sensors, e.g. MOS or CMOS sensors
    • H04N25/77—Pixel circuitry, e.g. memories, A/D converters, pixel amplifiers, shared circuits or shared components

Definitions

  • the present application relates to the field of imaging, and more particularly to an imaging element, an imaging device, and an image information processing method for acquiring edge information of an object.
  • a very important aspect of artificial intelligence is the automatic recognition of images or videos. How to quickly obtain accurate and refined image or video data for automatic identification of neural networks is a major challenge in the field of sensors.
  • Embodiments of the present application provide an imaging element, an imaging apparatus, and an image information processing method that can obtain edge information of an object directly, quickly, and accurately.
  • an imaging element comprising: a plurality of photosensitive devices arranged as pixels for receiving light from an imaged object and converting the light into a switching current; a current scaling unit for The conversion current is scaled to obtain a pixel processing current; the difference current acquisition unit is configured to subtract N times the average value of the pixel processing current of the surrounding pixels of the pixel for the pixel processing current for each pixel to obtain a difference current And an analog-to-digital conversion unit for performing analog-to-digital conversion based on the difference current to obtain edge information of the imaged object.
  • the photosensitive device is a photosensitive device for directly achieving photoelectric conversion or a photosensitive device for indirectly achieving photoelectric conversion.
  • the light from the imaged object is light emitted by the imaged object or light reflected by the imaged object.
  • imaging element further comprising at least one of a lens unit and a filter unit; and the plurality of photosensitive devices for receiving light from the imaged object transmitted through at least one of the lens unit and the filter unit .
  • the switching current includes at least one of a leakage current and a bias current.
  • the current conversion unit includes one of: a current mirror for scaling the conversion current to obtain a pixel processing current; and a conversion circuit for converting the conversion current into a voltage for scaling Obtaining a pixel processing current; and, a direct scaling circuit for directly scaling the conversion current to obtain a pixel processing current.
  • the conversion circuit is configured to: convert the conversion current into a voltage; and generate an amplification current at the voltage to obtain the pixel processing current.
  • the conversion circuit includes only active devices, or both active devices and passive devices.
  • the pixel processing current includes a leakage current and an error current generated by the current scaling unit and an added bias current.
  • the difference current acquisition unit is configured to obtain the difference current by one of the following methods: adding pixel processing currents of surrounding pixels of the pixel and dividing by the number of the surrounding pixels M Multiply by N; add 1/M times the pixel processing current of the surrounding pixels of the pixel and multiply by N; and add N/M times of the pixel processing current of the surrounding pixels of the pixel.
  • the N is an integer or a decimal.
  • imaging element further comprising: a difference current scaling unit for scaling the difference current to obtain a scaled difference current; and the analog to digital conversion unit configured to perform the scaled difference current Analog to digital conversion to obtain edge information of the imaged object.
  • the plurality of photosensitive devices are arranged in an arrangement similar to that of a photoreceptor of a biological retina.
  • an image forming apparatus comprising: an imaging element as described above; and a processing unit for performing image processing based on edge information of the imaged object.
  • an image information processing method comprising: receiving light from an imaged object; converting the light into a switching current by a plurality of photosensitive devices arranged as pixels; performing the switching current Scaling to obtain a pixel processing current; subtracting N times the average value of the pixel processing current of the surrounding pixels of the pixel for the pixel processing current for each pixel to obtain a difference current; and performing analog-to-digital conversion based on the difference current Obtaining edge information of the imaged object.
  • the photosensitive device is a photosensitive device for directly implementing photoelectric conversion or a photosensitive device for indirectly realizing photoelectric conversion.
  • the light from the imaged object is light emitted by the imaged object or light reflected by the imaged object.
  • receiving light from the imaged object includes receiving light from the imaged object transmitted through at least one of the lens unit and the filter unit.
  • the switching current includes at least one of a leakage current and a bias current.
  • scaling the conversion current to obtain a pixel processing current includes one of: scaling the conversion current by a current mirror to obtain a pixel processing current; converting the conversion current into a voltage Scaling is performed to obtain a pixel processing current; and the conversion current is directly scaled to obtain a pixel processing current.
  • scaling the conversion current into a voltage to obtain a pixel processing current includes converting the conversion current into a voltage through a conversion circuit; and generating an amplification current at the voltage to obtain a The pixel processing current is described.
  • the conversion circuit includes only active devices, or both active devices and passive devices.
  • the pixel processing current includes a leakage current and an error current generated by scaling the conversion current and an added bias current.
  • the difference current is obtained by one of the following methods: adding pixel processing currents of surrounding pixels of the pixel and dividing by the number M of the surrounding pixels and multiplying by N; The 1/M times of the pixel processing current of the surrounding pixels of the pixel are added and multiplied by N; and the N/M times of the pixel processing current of the surrounding pixels of the pixel are added.
  • the N is an integer or a decimal.
  • the edge information includes: performing analog-to-digital conversion on the scaled difference current to obtain edge information of the imaged object.
  • the plurality of photosensitive devices are arranged in an arrangement similar to the photoreceptor of the biological retina.
  • the edge information of the object can be obtained directly, quickly, and accurately.
  • FIG. 1 illustrates a schematic block diagram of an imaging element in accordance with an embodiment of the present application.
  • Figure 2 is a graphical representation of a comparison of human retinal receptive patterns with artificial retinal circuit models.
  • FIG. 3 illustrates a circuit schematic of each pixel unit in an imaging element in accordance with an embodiment of the present application.
  • FIG. 4 illustrates a schematic block diagram of an image forming apparatus according to an embodiment of the present application.
  • FIG. 5 illustrates a schematic flowchart of an image information processing method according to an embodiment of the present application.
  • Image sensors are generally classified into charge coupled device (CCD) sensors and complementary metal oxide semiconductor (CMOS) sensors.
  • CCD charge coupled device
  • CMOS complementary metal oxide semiconductor
  • the CCD sensor has the disadvantages of high power consumption, high defect rate, and is not easy to integrate with the processing circuit.
  • CMOS sensor has the advantages of low power consumption and easy integration with processing circuits, and is becoming more and more popular.
  • the basic working principle of the CMOS sensor is to reset the photodiode to a certain voltage Vrst and then start the exposure; the photodiode converts the received light into a current, and discharges the parasitic capacitance; after a period of exposure, the photodiode
  • CMOS sensor still has drawbacks.
  • image information is redundant, and the amount of data is large;
  • second it is easy to saturate under strong light;
  • third the shooting speed is greatly affected by the light intensity;
  • fourth because of the use of line scan sampling quantization, the shooting speed is difficult to increase.
  • the basic idea of the embodiments of the present application is to provide an imaging element, an imaging device, and an image information processing method, which can extract only the luminance spatial gradient signal of the object and directly operate the analog current of the photosensitive element. Therefore, by extracting only the luminance spatial gradient signal of the object, the data redundancy can be significantly reduced, thereby reducing the required data transmission bandwidth and improving the working efficiency of the system without degrading the performance such as the recognition rate. Further, by directly operating the analog current of the photosensitive element, the intermediate conversion to voltage is omitted, and the light intensity limitation is eliminated. In addition, since the sampling speed is independent of the light intensity, high-speed shooting can be maintained under strong light and low light. Moreover, by directly implementing the analog current sum calculation in an analog circuit, subsequent complicated digital circuit design is avoided, and the power consumption and area of the imaging element are reduced.
  • the imaging element, the imaging device, and the image information processing method according to the embodiments of the present application can be used for image capturing, video recording, and thus can be widely applied to the field of artificial intelligence, such as image recognition, video recognition, automatic driving, and drone control. Wait.
  • FIG. 1 illustrates a schematic block diagram of an imaging element in accordance with an embodiment of the present application.
  • an imaging element 100 includes: a plurality of photosensitive devices 110 arranged as pixels for receiving light from an imaged object and converting the light into a switching current; a current scaling unit 120, The scaling current generated by the photosensitive device 110 is scaled to obtain a pixel processing current; the difference current acquiring unit 130 is configured to subtract the pixel from the current processing unit for each pixel The pixel processing current of the surrounding pixels is N times the average value of the current to obtain a difference current; and the analog-to-digital conversion unit 140 is configured to perform analog-to-digital conversion based on the difference current obtained by the difference current acquiring unit 130 to obtain the imaged object. Edge information.
  • the imaging element 100 directly operates the analog current of the photosensitive element instead of being converted into a voltage and then operated so as not to be limited by the light intensity. Moreover, since the sampling speed is independent of the light intensity, high-speed shooting can be maintained under strong light and low light.
  • the difference current acquisition unit 130 may directly implement analog current sum calculation by an analog circuit, specifically, an analog current adder and an analog current subtractor. In this way, subsequent complicated digital circuit design is avoided, and the power consumption and area of the imaging element are correspondingly reduced.
  • the plurality of photosensitive devices 110 arranged as pixels realize parallel synchronous sampling of all pixels, which can significantly increase the number of frames.
  • the photosensitive device 110 may be a photosensitive device for directly implementing photoelectric conversion, such as a photodiode or a phototransistor.
  • the photosensitive device 110 may also be a photosensitive device for indirectly implementing photoelectric conversion, such as a photoresistor. It will be understood by those skilled in the art that in the implementation of the imaging element 100 according to an embodiment of the present application, a suitable type of photosensitive device can be selected according to specific requirements, such as actual process requirements.
  • the photosensitive device 110 receives light from an imaging object, where the light from the imaging object may be light emitted by the imaging object or light reflected by the imaging object. .
  • the light from the imaged object may be directly irradiated onto the photosensitive device 110 from the imaged object, or may be light irradiated onto the photosensitive device 110 via other optical devices such as lenses or filters.
  • the light from the imaged object may be visible light, or may be non-visible light such as infrared or ultraviolet light, or may be other rays.
  • At least one of a lens unit and a filter unit may be further included; and the plurality of photosensitive devices 110 are used for reception via the lens unit and the filter unit At least one transmitted light from the imaged object.
  • the photosensitive device 110 may introduce some other leakage current in addition to the current itself converted by the light.
  • some bias current can be added to the switching current to allow the components to operate in the proper operating range to reduce the difficulty of circuit design and the errors introduced by the process.
  • the switching current generated by the photosensitive device 110 includes a leakage current and/or a bias current.
  • the current scaling unit 120 is configured to scale the conversion current generated by the photosensitive device 110 to obtain a pixel processing current, and, in the imaging element 100 according to an embodiment of the present application, the current scaling unit 120 may have A variety of implementation forms.
  • the current scaling unit 120 can include a current mirror for scaling the conversion current to obtain a pixel processing current.
  • the current scaling unit 120 may include a conversion circuit for converting the converted current into a voltage, scaling the voltage, and then converting the scaled voltage into a current to obtain a pixel processing current.
  • the current scaling unit 120 may also include a direct scaling circuit for directly scaling the conversion current to obtain a pixel processing current.
  • the conversion current in the case that the conversion current is converted into a voltage by using a conversion circuit, the conversion current may be first converted into a voltage, and then an amplified current is generated by the converted voltage, and finally the pixel processing current is obtained.
  • the conversion circuit may include only active devices, and may include both active devices and passive devices.
  • the current scaling unit 120 is required to scale the conversion current generated by the photosensitive device 110. Specifically, in the case where the illuminance is very weak, the switching current generated by the photosensitive device 110 is too small, and is not easily used for effective processing. In this case, the current needs to be amplified to reduce the difficulty and complexity of the subsequent processing circuit; In the case where the illuminance is strong, the switching current generated by the photosensitive device 110 is too large, and at this time, it is necessary to reduce the collected current to reduce the power consumption and the like.
  • the influence of the illuminance of the light can be minimized by the current scaling unit 120.
  • the current scaling unit 120 also introduces a leakage current and an error current.
  • a bias current or a voltage may be further added to reduce the difficulty of circuit design and the error introduced by the process.
  • the pixel processing current includes a leakage current and an error current generated by the current scaling unit 120 and an added bias current.
  • the difference current acquisition unit 130 is configured to subtract N times the average value of the pixel processing current of the surrounding pixels of the pixel for each pixel to obtain a difference current. That is, the difference current acquisition unit 130 obtains the pixel processing current of the photosensitive device 110 corresponding to each pixel (that is, the converted current converted by the photosensitive device 110 into a converted current) and corresponds to The difference between the pixel processing currents of the photosensitive device 110 of the surrounding pixels of the pixel. And, the difference current is N times the average value of the pixel processing current of the photosensitive device 110 minus the pixel processing current of the photosensitive device around it, which is the current representation of the luminance spatial gradient information of the object.
  • the arrangement shape of the photosensitive device 110 may be set as needed, such as a triangle, a quadrangle, a pentagon, a hexagon, etc., or even only one or two of up and down or left and right may be included.
  • Adjacent photosensitive device Adjacent photosensitive device.
  • adjacent pixels of a certain pixel may be directly adjacent pixels or indirectly adjacent pixels with other pixels interposed therebetween, which may be defined according to the complexity of the circuit and the fineness of the edges.
  • M can be any number between 1 and 10000.
  • the N is an integer or a decimal.
  • the value of N is determined by pixel size and process characteristics, thereby reducing the complexity of subsequent circuits and improving image quality.
  • N can be any number between 0.01 and 10,000.
  • the difference current acquisition unit 130 can obtain the difference current in various manners. For example, the difference current acquisition unit 130 may first add pixel processing currents of surrounding pixels of the pixel, divide by the number M of the surrounding pixels to obtain the average value, and finally multiply N to obtain the Poor current. In addition, the difference current acquisition unit 130 may add 1/M times the pixel processing current of the surrounding pixels of the pixel to obtain the average value, and multiply by N to obtain the difference current. Moreover, the difference current acquisition unit 130 may directly add N/M times of the pixel processing current of the surrounding pixels of the pixel, thereby obtaining the difference current. In the specific implementation process, the corresponding analog circuit can be selected according to the actual needs of the design, thereby directly calculating the analog current.
  • the analog-to-digital conversion unit 140 directly performs current-mode analog-to-digital conversion on the difference current to obtain edge information of the imaged object, thereby realizing digitization of edge information of the object.
  • a differential current scaling unit may be further included for scaling the difference current to obtain a scaled difference current.
  • the analog-to-digital conversion unit 140 is configured to perform analog-to-digital conversion on the scaled difference current to obtain edge information of the imaged object.
  • the differential current needs to be amplified to reduce the difficulty and complexity of the subsequent processing circuit; and if the difference current is too large, it may be To cause unnecessary power consumption, the reduction process will be performed at this time.
  • the imaging element 100 can be applied to the field of artificial intelligence to recognize images and videos.
  • imaging element 100 in accordance with embodiments of the present application can be modeled in the retina model of the human eye.
  • Figure 2 is a graphical representation of a comparison of human retinal receptive patterns with artificial retinal circuit models.
  • the photodiode, the analog current adder, the current scaling element, and the analog current subtractor in the artificial imitation retinal circuit model correspond to photoreceptors, ganglion cells, bipolar cells, and horizontal cells in the human eye, respectively.
  • the imaging element 100 can be used not only to model the retina of the human eye but also to emulate the retina of other organisms.
  • the plurality of photosensitive devices 110 are arranged in an arrangement similar to the photoreceptor of the biological retina.
  • FIG. 3 illustrates a circuit schematic of each pixel unit in an imaging element in accordance with an embodiment of the present application.
  • the photosensitive element PD 0 corresponding to a certain pixel and the photosensitive elements PD 1 , PD 2 , . . . , PD n corresponding to the surrounding pixels thereof respectively obtain conversion currents I d0 , I d1 , I d2 , . .., I dn .
  • I dn are converted by the current amplifying circuit Amp and then summed by an analog current adder, for example, Then, the analog current subtractor is subtracted from the current I d0 to obtain a difference current ⁇ I, that is, a current expression of the luminance spatial gradient signal of the object. Then, it is digitized via a current-mode analog-to-digital converter ADC, and finally the edge information D 0 of the image is obtained.
  • FIG. 4 illustrates a schematic block diagram of an image forming apparatus according to an embodiment of the present application.
  • an imaging apparatus 200 includes: an imaging element 210 for obtaining edge information of the imaged object by receiving light from the imaged object and converting the light into a current; and processing The unit 220 is configured to perform image processing based on edge information of the imaged object obtained by the imaging element 210.
  • the processing unit 220 of the imaging apparatus 200 may perform image processing based on edge information of the imaged object, including but not limited to object detection, object recognition, object tracking, and the like.
  • the imaging element 210 includes: a plurality of photosensitive devices arranged as pixels for receiving light from the imaged object and converting the light into a switching current; and a current scaling unit for performing the conversion current Scaling to obtain a pixel processing current; a difference current obtaining unit for subtracting N times the average value of the pixel processing current of the surrounding pixels of the pixel for the pixel processing current for each pixel to obtain a difference current; and, the modulus And a conversion unit configured to perform analog-to-digital conversion based on the difference current to obtain edge information of the imaged object.
  • the photosensitive device is a photosensitive device for directly achieving photoelectric conversion or a photosensitive device for indirectly achieving photoelectric conversion.
  • the light from the imaged object is light emitted by the imaged object or light reflected by the imaged object.
  • the imaging element 210 further includes at least one of a lens unit and a filter unit; and the plurality of photosensitive devices are configured to receive via the lens unit and the filter unit At least one of the transmitted light from the imaged object.
  • the switching current includes at least one of a leakage current and a bias current.
  • the current conversion unit includes one of: a current mirror for scaling the conversion current to obtain a pixel processing current; and a conversion circuit for converting the conversion current Converting to voltage for scaling to obtain pixel processing current; and, direct scaling circuitry for directly scaling the conversion current to obtain pixel processing current.
  • the conversion circuit is configured to: convert the conversion current into a voltage; and generate an amplification current at the voltage to obtain the pixel processing current.
  • the conversion circuit includes only active devices, or both active devices and passive devices.
  • the pixel processing current includes a leakage current and an error current generated by the current scaling unit and an added bias current.
  • the difference current acquisition unit is configured to obtain the difference current by one of the following methods: adding pixel processing currents of surrounding pixels of the pixel and dividing by The number M of surrounding pixels is multiplied by N; the 1/M times of the pixel processing current of the surrounding pixels of the pixel are added and multiplied by N; and the pixel processing current of the surrounding pixels of the pixel is N/ Add M times.
  • the N is an integer or a decimal.
  • the imaging element 210 further includes: a difference current scaling unit that scales the difference current to obtain a scaled difference current; and the analog to digital conversion unit And performing analog-to-digital conversion on the scaled difference current to obtain edge information of the imaged object.
  • the plurality of photosensitive devices are arranged in an arrangement similar to that of a photoreceptor of a biological retina.
  • FIG. 5 illustrates a schematic flowchart of an image information processing method according to an embodiment of the present application.
  • the image information processing method includes: S310, receiving light from an imaged object; S320, converting the light into a converted current by a plurality of photosensitive devices arranged as pixels; S330, The conversion current is scaled to obtain a pixel processing current; S340, subtracting N times the average value of the pixel processing current of the surrounding pixels of the pixel for the pixel processing current for each pixel to obtain a difference current; and S350, based on The difference current is subjected to analog to digital conversion to obtain edge information of the imaged object.
  • the photosensitive device is a photosensitive device for directly implementing photoelectric conversion or a photosensitive device for indirectly realizing photoelectric conversion.
  • the light from the imaged object is light emitted by the imaged object or light reflected by the imaged object.
  • receiving light from the imaged object includes receiving light from the imaged object transmitted through at least one of the lens unit and the filter unit.
  • the switching current includes at least one of a leakage current and a bias current.
  • scaling the conversion current to obtain a pixel processing current includes one of: scaling the conversion current by a current mirror to obtain a pixel processing current; converting the conversion current into a voltage Scaling is performed to obtain a pixel processing current; and the conversion current is directly scaled to obtain a pixel processing current.
  • scaling the conversion current into a voltage to obtain a pixel processing current includes converting the conversion current into a voltage through a conversion circuit; and generating an amplification current at the voltage to obtain a The pixel processing current is described.
  • the conversion circuit includes only active devices, or both active devices and passive devices.
  • the pixel processing current includes a leakage current and an error current generated by scaling the conversion current and an added bias current.
  • the difference current is obtained by one of the following methods: adding pixel processing currents of surrounding pixels of the pixel and dividing by the number M of the surrounding pixels and multiplying by N; The 1/M times of the pixel processing current of the surrounding pixels of the pixel are added and multiplied by N; and the N/M times of the pixel processing current of the surrounding pixels of the pixel are added.
  • the N is an integer or a decimal.
  • the edge information includes: performing analog-to-digital conversion on the scaled difference current to obtain edge information of the imaged object.
  • the plurality of photosensitive devices are arranged in an arrangement similar to the photoreceptor of the biological retina.
  • the imaging element, the imaging apparatus, and the image information processing method according to the present application can extract only the luminance spatial gradient signal of the object, thereby significantly reducing data redundancy. Therefore, the required data transmission bandwidth is reduced and the operating efficiency of the system is improved without degrading the performance such as the recognition rate.
  • the imaging element, the imaging device, and the image information processing method according to the present application can directly operate the analog current of the photosensitive element instead of being converted into a voltage and then operated so as not to be limited by the light intensity. Moreover, since the sampling speed is independent of the light intensity, high-speed shooting can be maintained under strong light and low light.
  • the imaging element, the imaging device, and the image information processing method according to the present application can directly realize the summation of the analog current and the difference calculation by the analog circuit. In this way, subsequent complicated digital circuit design is avoided, and the power consumption and area of the imaging element are correspondingly reduced.
  • the imaging element, the imaging apparatus, and the image information processing method of the present application parallel simultaneous sampling of all pixels can be realized by a plurality of photosensitive devices arranged as pixels, and the number of frames can be remarkably increased.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Transforming Light Signals Into Electric Signals (AREA)
  • Studio Devices (AREA)

Abstract

本申请涉及一种成像元件，包括：布置为像素的多个感光器件，用于接收来自成像物体的光线并将所述光线转换为转换电流；电流缩放单元，用于对所述转换电流进行缩放以获得像素处理电流；差电流获取单元，用于将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及，模数转换单元，用于基于所述差电流进行模数转换以获得所述成像物体的边沿信息。这样，可以直接、快速和准确地获得物体的边沿信息。

Description

成像元件、成像设备和图像信息处理方法 技术领域
本申请涉及成像领域，更为具体地说，涉及用于获取物体的边沿信息的成像元件、成像设备和图像信息处理方法。
背景技术
当今时代，人工智能澎湃发展，从各个方面影响着人们的生产和生活，同时也推动着世界的发展和进步。
而人工智能中一个很重要的方面，就是图像或视频的自动识别。如何快速获得准确而精炼的图像或视频数据，以便于神经网络进行自动识别，是传感器领域的一大挑战。
因此，需要用于获取图像信息的改进的成像元件、成像设备和图像信息处理方法。
发明内容
为了解决上述技术问题，提出了本申请。本申请的实施例提供了成像元件、成像设备和图像信息处理方法，其可以直接、快速和准确地获得物体的边沿信息。
根据本申请的一方面，提供了一种成像元件，包括：布置为像素的多个感光器件，用于接收来自成像物体的光线并将所述光线转换为转换电流；电流缩放单元，用于对所述转换电流进行缩放以获得像素处理电流；差电流获取单元，用于将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及，模数转换单元，用于基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
在上述成像元件中，所述感光器件是用于直接实现光电转换的感光器件或者用于间接实现光电转换的感光器件。
在上述成像元件中，所述来自成像物体的光线是所述成像物体发出的光线或者所述成像物体反射的光线。
在上述成像元件中，进一步包括透镜单元和滤光单元中的至少一个；以及，所述多个感光器件用于接收经由所述透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
在上述成像元件中，所述转换电流包括漏电流和偏置电流中的至少一个。
在上述成像元件中，所述电流转换单元包括以下的其中之一：电流镜，用于缩放所述转换电流以获得像素处理电流；转换电路，用于将所述转换电流转换为电压而进行缩放以获得像素处理电流；和，直接缩放电路，用于直接对所述转换电流进行缩放以获得像素处理电流。
在上述成像元件中，所述转换电路用于：将所述转换电流转换为电压；以及，以所述电压生成放大电流以获得所述像素处理电流。
在上述成像元件中，所述转换电路仅包括有源器件，或者包括有源器件和无源器件两者。
在上述成像元件中，所述像素处理电流包括所述电流缩放单元生成的漏电流和误差电流以及加入的偏置电流。
在上述成像元件中，所述差电流获取单元用于通过以下的其中一种方式获得所述差电流：将所述像素的周围像素的像素处理电流相加再除以所述周围像素的数目M再乘以N；将所述像素的周围像素的像素处理电流的1/M倍相加再乘以N；以及，将所述像素的周围像素的像素处理电流的N/M倍相加。
在上述成像元件中，所述N是整数或者小数。
在上述成像元件中，进一步包括：差电流缩放单元，用于对所述差电流进行缩放以获得缩放后的差电流；以及，所述模数转换单元用于对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
在上述成像元件中，所述多个感光器件以与生物视网膜的光感受体类似的布置形式布置。
根据本申请的另一方面，提供了一种成像设备，包括：如上所述的成像元件；以及处理单元，用于基于所述成像物体的边沿信息进行图像处理。
根据本申请的再一方面，提供了一种图像信息处理方法，包括：接收来自成像物体的光线；以布置为像素的多个感光器件将所述光线转换为转换电流；对所述转换电流进行缩放以获得像素处理电流；将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得 差电流；以及，基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
在上述图像信息处理方法中，所述感光器件是用于直接实现光电转换的感光器件或者用于间接实现光电转换的感光器件。
在上述图像信息处理方法中，所述来自成像物体的光线是所述成像物体发出的光线或者所述成像物体反射的光线。
在上述图像信息处理方法中，接收来自成像物体的光线包括：接收经由透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
在上述图像信息处理方法中，所述转换电流包括漏电流和偏置电流中的至少一个。
在上述图像信息处理方法中，对所述转换电流进行缩放以获得像素处理电流包括以下的其中之一：通过电流镜缩放所述转换电流以获得像素处理电流；将所述转换电流转换为电压而进行缩放以获得像素处理电流；和，直接对所述转换电流进行缩放以获得像素处理电流。
在上述图像信息处理方法中，将所述转换电流转换为电压而进行缩放以获得像素处理电流包括：通过转换电路将所述转换电流转换为电压；以及，以所述电压生成放大电流以获得所述像素处理电流。
在上述图像信息处理方法中，所述转换电路仅包括有源器件，或者包括有源器件和无源器件两者。
在上述图像信息处理方法中，所述像素处理电流包括对所述转换电流进行缩放而生成的漏电流和误差电流以及加入的偏置电流。
在上述图像信息处理方法中，通过以下的其中一种方式获得所述差电流：将所述像素的周围像素的像素处理电流相加再除以所述周围像素的数目M再乘以N；将所述像素的周围像素的像素处理电流的1/M倍相加再乘以N；以及，将所述像素的周围像素的像素处理电流的N/M倍相加。
在上述图像信息处理方法中，所述N是整数或者小数。
在上述图像信息处理方法中，在获得所述差电流之后进一步包括：对所述差电流进行缩放以获得缩放后的差电流；以及，基于所述差电流进行模数转换以获得所述成像物体的边沿信息包括：对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
在上述图像信息处理方法中，所述多个感光器件以与生物视网膜的光感 受体类似的布置形式布置。
与现有技术相比，采用根据本申请实施例的成像元件、成像设备和图像信息处理方法，可以直接、快速和准确地获得物体的边沿信息。
附图说明
通过阅读下文优选的具体实施方式中的详细描述，本申请各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。说明书附图仅用于示出优选实施方式的目的，而并不认为是对本申请的限制。显而易见地，下面描述的附图仅仅是本申请的一些实施例，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。而且在整个附图中，用相同的附图标记表示相同的部件。
图1图示了根据本申请实施例的成像元件的示意性框图。
图2图示了人眼视网膜感受模式与人工仿视网膜电路模型的对比示意图。
图3图示了根据本申请实施例的成像元件中每个像素单元的电路示意图。
图4图示了根据本申请实施例的成像设备的示意性框图。
图5图示了根据本申请实施例的图像信息处理方法的示意性流程图。
具体实施方式
下面，将参考附图详细地描述根据本申请的示例实施例。显然，所描述的实施例仅仅是本申请的一部分实施例，而不是本申请的全部实施例，应理解，本申请不受这里描述的示例实施例的限制。
申请概述
如上所述，在图像或者视频识别中，需要对物体进行成像，并获得相应的图像信息。
图像传感器一般分为电荷耦合器件(CCD)传感器和互补金属氧化物半导体(CMOS)传感器。其中，CCD传感器具有功耗高，缺陷率高，不易于与处理电路集成等缺点。而CMOS传感器具有功耗低，易与处理电路集成等优点，越来越流行。
CMOS传感器的基本工作原理是：将感光二极管复位至某一电压Vrst，然后开始曝光；光电二极管将接收的光转换成电流，对自身寄生的电容进行 放电；等一段时间曝光结束，此时感光二极管的电压为Vo；计算ΔV＝Vrst-Vo，将ΔV量化，最终得到像素的亮度。
但是，上述CMOS传感器仍然具有缺陷。第一，图像信息冗余，数据量大；第二，强光下易饱和；第三，拍摄速度受光照强度影响很大；第四，由于使用行扫描采样量化，所以拍摄速度难以提高。
针对上述技术问题，本申请实施例的基本构思是提出一种成像元件、成像设备和图像信息处理方法，可以仅提取物体的亮度空间梯度信号，并直接对感光元件的模拟电流进行操作。因此，通过仅提取物体的亮度空间梯度信号，可以显著降低数据冗余，从而在不降低识别率等性能的同时，相应降低了所需的数据传输带宽，提高了系统的工作效率。并且，通过直接对感光元件的模拟电流进行操作，而省略了中间转换成电压的步骤，消除了光强限制。另外，由于采样速度与光强无关，在强光弱光下均可保持高速拍摄。而且，通过以模拟电路直接实现模拟电流的和差计算，避免了后续复杂的数字电路设计，降低了成像元件的功耗和面积。
这里，根据本申请实施例的成像元件、成像设备和图像信息处理方法可用于图像拍摄，视频录制，从而可以广泛地应用于人工智能领域，比如图像识别，视频识别，自动驾驶，无人机控制等。
在介绍了本申请的基本原理之后，下面将参考附图来具体介绍本申请的各种非限制性实施例。
示例性成像元件
图1图示了根据本申请实施例的成像元件的示意性框图。
如图1所示，根据本申请实施例的成像元件100包括：布置为像素的多个感光器件110，用于接收来自成像物体的光线并将所述光线转换为转换电流；电流缩放单元120，用于对所述感光器件110生成的所述转换电流进行缩放以获得像素处理电流；差电流获取单元130，用于将所述电流缩放单元120针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及，模数转换单元140，用于基于所述差电流获取单元130获得的差电流进行模数转换以获得所述成像物体的边沿信息。
因此，根据本申请实施例的成像元件100可以仅提取物体的亮度空间梯 度信号，从而显著降低数据冗余。因为在图像或者视频识别中，物体的边沿信息剔除了不相关的信息，而保留了图像重要的结构属性。因此，通过获得物体的边沿信息，可以在不降低识别率等性能的情况下，降低所需的数据传输带宽，并提高系统的工作效率。
另外，根据本申请实施例的成像元件100直接对感光元件的模拟电流进行操作，而不是转换成电压再进行操作，使得不受光强限制。而且，由于采样速度与光强无关，在强光弱光下均可保持高速拍摄。
在根据本申请实施例的成像元件100中，所述差电流获取单元130可以通过模拟电路，具体来说，模拟电流加法器和模拟电流减法器直接实现模拟电流的和差计算。这样，避免了后续复杂的数字电路设计，也相应地降低了成像元件的功耗和面积。
而且，在根据本申请实施例的成像元件100中，布置为像素的多个感光器件110实现了所有像素的并行同步采样，这可以显著提高帧数。
这里，感光器件110可以是用于直接实现光电转换的感光器件，比如光电二极管或者光电三极管。或者，感光器件110也可以是用于间接实现光电转换的感光器件，比如光敏电阻。本领域技术人员可以理解，在根据本申请实施例的成像元件100的实现过程中，可以根据具体要求，比如实际工艺要求选择适当类型的感光器件。
在根据本申请实施例的成像元件100中，感光器件110接收来自成像物体的光线，这里，所述来自成像物体的光线可以是所述成像物体发出的光线，或者是所述成像物体反射的光线。
并且，所述来自成像物体的光线可以从成像物体直接照射在感光器件110上的光线，也可以是经由其它光学器件，例如透镜或者滤光片等照射在感光器件110上的光线。并且，这里，所述来自成像物体的光线可以是可见光，也可以是红外或者紫外等非可见光，也可以是其它射线。
因此，在根据本申请实施例的成像元件100中，可以进一步包括透镜单元和滤光单元中的至少一个；以及，所述多个感光器件110用于接收经由所述透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
所述感光器件110在将所述光线转换为转换电流的过程中，除了由所述光线转换而来的电流本身外，还可能引入一些其它的漏电流。另外，可以在转换电流中加入一些偏置电流，使得元器件工作在合适的工作区间，以减小 电路设计的难度和工艺引入的误差。
因此，在根据本申请实施例的成像元件100中，所述感光器件110生成的转换电流包括漏电流和/或偏置电流。
所述电流缩放单元120用于对所述感光器件110生成的所述转换电流进行缩放以获得像素处理电流，并且，在根据本申请实施例的成像元件100中，所述电流缩放单元120可以有多种实现形式。
具体来说，所述电流缩放单元120可以包括电流镜，以用于对所述转换电流进行缩放以获得像素处理电流。此外，所述电流缩放单元120可以包括转换电路，用于将所述转换电流转换为电压，再对所述电压进行缩放，之后再将缩放后的电压转换为电流以获得像素处理电流。当然，所述电流缩放单元120也可以包括直接缩放电路，用于直接对所述转换电流进行缩放以获得像素处理电流。
其中，在采用转换电路将所述转换电流转换为电压的情况下，可以首先将所述转换电流转换为电压，再通过转换得到的电压生成放大电流，最终获得所述像素处理电流。并且，所述转换电路可以仅包括有源器件，也可以既包括有源器件又包括无源器件。
由于在不同照度条件下，感光器件110的电流变化范围非常大，因此在根据本申请实施例的成像元件100中，需要电流缩放单元120来对感光器件110生成的转换电流进行缩放。具体来说，在照度很弱的情况下，感光器件110生成的转换电流过小，不容易直接用来进行有效处理，此时需要将电流放大以降低后继处理电路的难度和复杂度等；另外，在照度很强的情况下，感光器件110生成的转换电流过大，此时又需要缩小采集到的电流以降低功耗等。
因此，通过所述电流缩放单元120，可以尽量减小光的照度带来的影响。
并且，在放大或者缩小所述转换电流以获得所述像素处理电流的过程中，所述电流缩放单元120也会引入漏电流和误差电流。相应地，在获得的所述像素处理电流中，也可以再加入偏置电流或者电压，以减小电路设计的难度和工艺引入的误差。
因此，在根据本申请实施例的成像元件100中，所述像素处理电流包括所述电流缩放单元120生成的漏电流和误差电流以及加入的偏置电流。
所述差电流获取单元130用于针对每一像素，将其像素处理电流减去所 述像素的周围像素的像素处理电流的平均值的N倍以获得差电流。也就是说，所述差电流获取单元130获得对应于每一像素的感光器件110的像素处理电流(即，所述感光器件110将光线转换为的转换电流经缩放后获得的)和对应于所述像素的周围像素的感光器件110的像素处理电流之间的差电流。并且，所述差电流是所述感光器件110的像素处理电流减去其周围的感光器件的像素处理电流的平均值的N倍，其是物体的亮度空间梯度信息的电流表现。
在根据本申请实施例的成像元件100中，可以按照需要设置感光器件110的布置形状，例如三角形，四边形，五边形，六边形等，甚至也可以仅包含上下或者左右的一个或者两个相邻感光器件。另外，关于某个像素的相邻像素，既可以是直接相邻的像素，也可以是中间间隔有其它像素的间接相邻的像素，这可以根据电路的复杂度以及边沿的精细度来定义。具体来说，M可以是1到10000之间的任意数。
其中，所述N是整数或者小数。这里，通过像素大小和工艺特点来确定N的值，从而降低后续电路的复杂度和改善图像质量。具体来说，N可以是0.01到10000之间的任意数。
具体来说，所述差电流获取单元130可以通过多种方式获得所述差电流。例如，所述差电流获取单元130可以先将所述像素的周围像素的像素处理电流相加，再除以所述周围像素的数目M以获得所述平均值，最后乘以N以获得所述差电流。另外，所述差电流获取单元130可以将所述像素的周围像素的像素处理电流的1/M倍相加以获得所述平均值，再乘以N以获得所述差电流。并且，所述差电流获取单元130还可以直接将所述像素的周围像素的像素处理电流的N/M倍相加，从而获得所述差电流。在具体的实现过程中，可以根据设计的实际需要选取相应的模拟电路，从而直接对模拟电流进行运算。
所述模数转换单元140对所述差电流直接进行电流型模数转换，以获得所述成像物体的边沿信息，从而实现了物体的边沿信息数字化。
另外，可选地，在根据本申请实施例的成像元件100中，可以进一步包括差电流缩放单元，用于对所述差电流进行缩放以获得缩放后的差电流。并且，所述模数转换单元140用于对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
与上述的转换电流类似，如果差电流过小，不容易直接用来进行有效处 理，此时需要将差电流放大以降低后继处理电路的难度和复杂度等；而如果差电流过大，可能会造成不必要的功耗，此时就会进行缩小处理。
如上所述，根据本申请实施例的成像元件100可以应用于人工智能领域，以对图像和视频进行识别。因此，在一个示例中，根据本申请实施例的成像元件100可以仿照人眼的视网膜模型设置。
图2图示了人眼视网膜感受模式与人工仿视网膜电路模型的对比示意图。如图2所示，人工仿视网膜电路模型中的感光二极管、模拟电流加法器、电流缩放元件和模拟电流减法器分别对应于人眼中的光感受体、节细胞、双极细胞和水平细胞。
并且，本领域技术人员可以理解，根据本申请实施例的成像元件100不仅可用于仿照人眼视网膜，也可以用于仿照其它生物的视网膜。
因此，在根据本申请实施例的成像元件100中，所述多个感光器件110以与生物视网膜的光感受体类似的布置形式布置。
如上所述，为了直接、快速和准确地获得物体的边沿信息，根据本申请实施例的成像元件100以模拟电路实现。图3图示了根据本申请实施例的成像元件中每个像素单元的电路示意图。
如图3所示，对应于某像素的感光元件PD 0及对应于其周围像素的感光元件PD 1、PD 2、...、PD n分别获得转换电流I d0、I d1、I d2、...、I dn。其中，转换电流I d1、I d2、...、I dn经由电流放大电路Amp转换后经由模拟电流加法器求和，例如得到
Figure PCTCN2018115225-appb-000001
再经由模拟电流减法器与电流I d0相减，从而得到差电流ΔI，即物体的亮度空间梯度信号的电流表现。然后，经由电流型模数转换器ADC数字化，最终得到图像的边沿信息D 0。
示例性成像设备
图4图示了根据本申请实施例的成像设备的示意性框图。
如图4所示，根据本申请实施例的成像设备200包括：成像元件210，用于通过接收来自成像物体的光线并将所述光线转换为电流而获得所述成像物体的边沿信息；以及处理单元220，用于基于所述成像元件210获得的所述成像物体的边沿信息进行图像处理。
这里，根据本申请实施例的成像设备200的处理单元220可以基于所述 成像物体的边沿信息进行图像处理，包括但不限于对象检测、对象识别、对象跟踪等。
具体来说，所述成像元件210包括：布置为像素的多个感光器件，用于接收来自成像物体的光线并将所述光线转换为转换电流；电流缩放单元，用于对所述转换电流进行缩放以获得像素处理电流；差电流获取单元，用于将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及，模数转换单元，用于基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
在一个示例中，在上述成像设备200中，所述感光器件是用于直接实现光电转换的感光器件或者用于间接实现光电转换的感光器件。
在一个示例中，在上述成像设备200中，所述来自成像物体的光线是所述成像物体发出的光线或者所述成像物体反射的光线。
在一个示例中，在上述成像设备200中，所述成像元件210进一步包括透镜单元和滤光单元中的至少一个；以及，所述多个感光器件用于接收经由所述透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
在一个示例中，在上述成像设备200中，所述转换电流包括漏电流和偏置电流中的至少一个。
在一个示例中，在上述成像设备200中，所述电流转换单元包括以下的其中之一：电流镜，用于缩放所述转换电流以获得像素处理电流；转换电路，用于将所述转换电流转换为电压而进行缩放以获得像素处理电流；和，直接缩放电路，用于直接对所述转换电流进行缩放以获得像素处理电流。
在一个示例中，在上述成像设备200中，所述转换电路用于：将所述转换电流转换为电压；以及，以所述电压生成放大电流以获得所述像素处理电流。
在一个示例中，在上述成像设备200中，所述转换电路仅包括有源器件，或者包括有源器件和无源器件两者。
在一个示例中，在上述成像设备200中，所述像素处理电流包括所述电流缩放单元生成的漏电流和误差电流以及加入的偏置电流。
在一个示例中，在上述成像设备200中，所述差电流获取单元用于通过以下的其中一种方式获得所述差电流：将所述像素的周围像素的像素处理电流相加再除以所述周围像素的数目M再乘以N；将所述像素的周围像素的 像素处理电流的1/M倍相加再乘以N；以及，将所述像素的周围像素的像素处理电流的N/M倍相加。
在一个示例中，在上述成像设备200中，所述N是整数或者小数。
在一个示例中，在上述成像设备200中，所述成像元件210进一步包括：差电流缩放单元，用于对所述差电流进行缩放以获得缩放后的差电流；以及，所述模数转换单元用于对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
在一个示例中，在上述成像设备200中，所述多个感光器件以与生物视网膜的光感受体类似的布置形式布置。
这里，本领域技术人员可以理解，根据本申请实施例的成像设备的其它细节与之前关于根据本申请实施例的成像元件所描述的相应细节完全相同，这里为了避免冗余便不再赘述。
示例性图像信息处理方法
图5图示了根据本申请实施例的图像信息处理方法的示意性流程图。
如图5所示，根据本申请实施例的图像信息处理方法包括：S310，接收来自成像物体的光线；S320，以布置为像素的多个感光器件将所述光线转换为转换电流；S330，对所述转换电流进行缩放以获得像素处理电流；S340，将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及S350，基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
在上述图像信息处理方法中，所述感光器件是用于直接实现光电转换的感光器件或者用于间接实现光电转换的感光器件。
在上述图像信息处理方法中，所述来自成像物体的光线是所述成像物体发出的光线或者所述成像物体反射的光线。
在上述图像信息处理方法中，接收来自成像物体的光线包括：接收经由透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
在上述图像信息处理方法中，所述转换电流包括漏电流和偏置电流中的至少一个。
在上述图像信息处理方法中，对所述转换电流进行缩放以获得像素处理电流包括以下的其中之一：通过电流镜缩放所述转换电流以获得像素处理电 流；将所述转换电流转换为电压而进行缩放以获得像素处理电流；和，直接对所述转换电流进行缩放以获得像素处理电流。
在上述图像信息处理方法中，将所述转换电流转换为电压而进行缩放以获得像素处理电流包括：通过转换电路将所述转换电流转换为电压；以及，以所述电压生成放大电流以获得所述像素处理电流。
在上述图像信息处理方法中，所述转换电路仅包括有源器件，或者包括有源器件和无源器件两者。
在上述图像信息处理方法中，所述像素处理电流包括对所述转换电流进行缩放而生成的漏电流和误差电流以及加入的偏置电流。
在上述图像信息处理方法中，通过以下的其中一种方式获得所述差电流：将所述像素的周围像素的像素处理电流相加再除以所述周围像素的数目M再乘以N；将所述像素的周围像素的像素处理电流的1/M倍相加再乘以N；以及，将所述像素的周围像素的像素处理电流的N/M倍相加。
在上述图像信息处理方法中，所述N是整数或者小数。
在上述图像信息处理方法中，在获得所述差电流之后进一步包括：对所述差电流进行缩放以获得缩放后的差电流；以及，基于所述差电流进行模数转换以获得所述成像物体的边沿信息包括：对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
在上述图像信息处理方法中，所述多个感光器件以与生物视网膜的光感受体类似的布置形式布置。
这里，本领域技术人员可以理解，根据本申请实施例的图像信息处理方法的其它细节与之前关于根据本申请实施例的成像元件所描述的相应细节完全相同，这里为了避免冗余便不再赘述。
根据本申请的成像元件、成像设备和图像信息处理方法可以仅提取物体的亮度空间梯度信号，从而显著降低数据冗余。因此，在不降低识别率等性能的情况下，降低所需的数据传输带宽，并提高系统的工作效率。
根据本申请的成像元件、成像设备和图像信息处理方法可以直接对感光元件的模拟电流进行操作，而不是转换成电压再进行操作，使得不受光强限制。而且，由于采样速度与光强无关，在强光弱光下均可保持高速拍摄。
根据本申请的成像元件、成像设备和图像信息处理方法可以通过模拟电 路直接实现模拟电流的和差计算。这样，避免了后续复杂的数字电路设计，也相应地降低了成像元件的功耗和面积。
根据本申请的成像元件、成像设备和图像信息处理方法可以通过布置为像素的多个感光器件实现所有像素的并行同步采样，可以显著提高帧数。
以上结合具体实施例描述了本申请的基本原理，但是，需要指出的是，在本申请中提及的优点、优势、效果等仅是示例而非限制，不能认为这些优点、优势、效果等是本申请的各个实施例必须具备的。另外，上述公开的具体细节仅是为了示例的作用和便于理解的作用，而非限制，上述细节并不限制本申请为必须采用上述具体的细节来实现。
本申请中涉及的器件、装置、设备、系统的方框图仅作为例示性的例子并且不意图要求或暗示必须按照方框图示出的方式进行连接、布置、配置。如本领域技术人员将认识到的，可以按任意方式连接、布置、配置这些器件、装置、设备、系统。诸如“包括”、“包含”、“具有”等等的词语是开放性词汇，指“包括但不限于”，且可与其互换使用。这里所使用的词汇“或”和“和”指词汇“和/或”，且可与其互换使用，除非上下文明确指示不是如此。这里所使用的词汇“诸如”指词组“诸如但不限于”，且可与其互换使用。
还需要指出的是，在本申请的装置、设备和方法中，各部件或各步骤是可以分解和/或重新组合的。这些分解和/或重新组合应视为本申请的等效方案。
提供所公开的方面的以上描述以使本领域的任何技术人员能够做出或者使用本申请。对这些方面的各种修改对于本领域技术人员而言是非常显而易见的，并且在此定义的一般原理可以应用于其他方面而不脱离本申请的范围。因此，本申请不意图被限制到在此示出的方面，而是按照与在此公开的原理和新颖的特征一致的最宽范围。
为了例示和描述的目的已经给出了以上描述。此外，此描述不意图将本申请的实施例限制到在此公开的形式。尽管以上已经讨论了多个示例方面和实施例，但是本领域技术人员将认识到其某些变型、修改、改变、添加和子组合。

Claims (15)

  1. 一种成像元件，包括：
    布置为像素的多个感光器件，用于接收来自成像物体的光线并将所述光线转换为转换电流；
    电流缩放单元，用于对所述转换电流进行缩放以获得像素处理电流；
    差电流获取单元，用于将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及
    模数转换单元，用于基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
  2. 根据权利要求1所述的成像元件，其中，所述感光器件是用于直接实现光电转换的感光器件或者用于间接实现光电转换的感光器件。
  3. 根据权利要求1所述的成像元件，其中，所述来自成像物体的光线是所述成像物体发出的光线或者所述成像物体反射的光线。
  4. 根据权利要求1所述的成像元件，进一步包括透镜单元和滤光单元中的至少一个；以及
    所述多个感光器件用于接收经由所述透镜单元和滤光单元中的至少一个透射的来自成像物体的光线。
  5. 根据权利要求1所述的成像元件，其中，所述转换电流包括漏电流和偏置电流中的至少一个。
  6. 根据权利要求1所述的成像元件，其中，所述电流缩放单元包括以下的其中之一：
    电流镜，用于缩放所述转换电流以获得像素处理电流；
    转换电路，用于将所述转换电流转换为电压而进行缩放以获得像素处理电流；和
    直接缩放电路，用于直接对所述转换电流进行缩放以获得像素处理电流。
  7. 根据权利要求6所述的成像元件，其中，所述转换电路用于：
    将所述转换电流转换为电压；以及
    以所述电压生成放大电流以获得所述像素处理电流。
  8. 根据权利要求7所述的成像元件，其中，所述转换电路仅包括有源器件，或者包括有源器件和无源器件两者。
  9. 根据权利要求1所述的成像元件，其中，所述像素处理电流包括所述电流缩放单元生成的漏电流和误差电流以及加入的偏置电流。
  10. 根据权利要求1所述的成像元件，其中，所述差电流获取单元用于通过以下的其中一种方式获得所述差电流：
    将所述像素的周围像素的像素处理电流相加再除以所述周围像素的数目M再乘以N；
    将所述像素的周围像素的像素处理电流的1/M倍相加再乘以N；以及
    将所述像素的周围像素的像素处理电流的N/M倍相加。
  11. 根据权利要求1所述的成像元件，其中，所述N是整数或者小数。
  12. 根据权利要求1所述的成像元件，进一步包括：
    差电流缩放单元，用于对所述差电流进行缩放以获得缩放后的差电流；以及
    所述模数转换单元用于对所述缩放后的差电流进行模数转换以获得所述成像物体的边沿信息。
  13. 根据权利要求1所述的成像元件，其中，所述多个感光器件以与生物视网膜的光感受体类似的布置形式布置。
  14. 一种成像设备，包括：
    如权利要求1-13中任一项所述的成像元件；以及
    处理单元，用于基于所述成像物体的边沿信息进行图像处理。
  15. 一种图像信息处理方法，包括：
    接收来自成像物体的光线；
    以布置为像素的多个感光器件将所述光线转换为转换电流；
    对所述转换电流进行缩放以获得像素处理电流；
    将针对每一像素的像素处理电流减去所述像素的周围像素的像素处理电流的平均值的N倍以获得差电流；以及
    基于所述差电流进行模数转换以获得所述成像物体的边沿信息。
PCT/CN2018/115225 2018-01-26 2018-11-13 成像元件、成像设备和图像信息处理方法 WO2019144678A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2020522333A JP6918390B2 (ja) 2018-01-26 2018-11-13 撮像素子、撮像機器及び画像情報処理方法

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810079401.9A CN108282623B (zh) 2018-01-26 2018-01-26 成像元件、成像设备和图像信息处理方法
CN201810079401.9 2018-01-26

Publications (1)

Publication Number Publication Date
WO2019144678A1 true WO2019144678A1 (zh) 2019-08-01

Family

ID=62805430

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/115225 WO2019144678A1 (zh) 2018-01-26 2018-11-13 成像元件、成像设备和图像信息处理方法

Country Status (3)

Country Link
JP (1) JP6918390B2 (zh)
CN (1) CN108282623B (zh)
WO (1) WO2019144678A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112600996A (zh) * 2020-12-03 2021-04-02 清华大学 紫外仿生视觉传感器

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108282623B (zh) * 2018-01-26 2019-10-11 北京灵汐科技有限公司 成像元件、成像设备和图像信息处理方法
CN111083405B (zh) * 2019-12-24 2021-06-04 清华大学 双模态仿生视觉传感器像素读出系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6628330B1 (en) * 1999-09-01 2003-09-30 Neomagic Corp. Color interpolator and horizontal/vertical edge enhancer using two line buffer and alternating even/odd filters for digital camera
US20090161756A1 (en) * 2007-12-19 2009-06-25 Micron Technology, Inc. Method and apparatus for motion adaptive pre-filtering
CN103262524A (zh) * 2011-06-09 2013-08-21 郑苍隆 自动聚焦图像系统
CN107295236A (zh) * 2017-08-11 2017-10-24 深圳市唯特视科技有限公司 一种基于飞行时间传感器的快照差分成像方法
CN108282623A (zh) * 2018-01-26 2018-07-13 北京灵汐科技有限公司 成像元件、成像设备和图像信息处理方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08275059A (ja) * 1995-03-31 1996-10-18 Mitsubishi Electric Corp 受光素子回路、受光素子回路アレイ及び受光素子回路の構成方法
JPH11225289A (ja) * 1997-11-04 1999-08-17 Nikon Corp エッジ検出用固体撮像装置、並びに固体撮像装置の駆動によるエッジ検出方法
JP5728821B2 (ja) * 2010-04-12 2015-06-03 株式会社ニコン 撮像装置および電子カメラ

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6628330B1 (en) * 1999-09-01 2003-09-30 Neomagic Corp. Color interpolator and horizontal/vertical edge enhancer using two line buffer and alternating even/odd filters for digital camera
US20090161756A1 (en) * 2007-12-19 2009-06-25 Micron Technology, Inc. Method and apparatus for motion adaptive pre-filtering
CN103262524A (zh) * 2011-06-09 2013-08-21 郑苍隆 自动聚焦图像系统
CN107295236A (zh) * 2017-08-11 2017-10-24 深圳市唯特视科技有限公司 一种基于飞行时间传感器的快照差分成像方法
CN108282623A (zh) * 2018-01-26 2018-07-13 北京灵汐科技有限公司 成像元件、成像设备和图像信息处理方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112600996A (zh) * 2020-12-03 2021-04-02 清华大学 紫外仿生视觉传感器
CN112600996B (zh) * 2020-12-03 2022-12-09 清华大学 紫外仿生视觉传感器

Also Published As

Publication number Publication date
CN108282623A (zh) 2018-07-13
CN108282623B (zh) 2019-10-11
JP2020537456A (ja) 2020-12-17
JP6918390B2 (ja) 2021-08-11

Similar Documents

Publication Publication Date Title
TWI414179B (zh) Shooting parameter adjustment method for face detection and image capturing device for face detection
WO2019144678A1 (zh) 成像元件、成像设备和图像信息处理方法
CN107770521B (zh) 相机拍摄模式调整方法
KR101822661B1 (ko) 비전 인식 장치 및 방법
JP7331201B2 (ja) 撮像装置、撮像システム、車両走行制御システム、及び画像処理装置
CN107071234A (zh) 一种镜头阴影校正方法及装置
TWI615029B (zh) 圖像感測器及終端與成像方法
US10574872B2 (en) Methods and apparatus for single-chip multispectral object detection
US20150109515A1 (en) Image pickup apparatus, image pickup system, method of controlling image pickup apparatus, and non-transitory computer-readable storage medium
TWI554104B (zh) 影像資料匯集高動態範圍成像系統及相關方法
JP2009081714A (ja) 撮像装置および撮像装置における顔領域決定方法
WO2017101864A1 (zh) 图像传感器、控制方法和电子装置
JP4523629B2 (ja) 撮像装置
JP5127628B2 (ja) 画像処理装置および画像処理方法
JP7164579B2 (ja) 画像センサー及び電子デバイス
CN107358591A (zh) 一种基于rgb交叉通道先验的简单透镜成像色差校正方法
CN114424522A (zh) 图像处理装置、电子设备、图像处理方法与程序
CN108076284A (zh) 相机拍摄模式调整系统
WO2023181630A1 (ja) 画像処理装置、および撮像システム
CN112669817B (zh) 语言识别方法、装置及电子设备
WO2024093240A1 (zh) 亮度变化检测装置及方法、电子设备
Kong et al. A Bio-Inspired 128× 128 Complementary Metal–Oxide–Semiconductor Vision Chip for Edge Detection with Signal Processing Circuit Separated from Photo-Sensing Circuit
TWI388201B (zh) 利用遮罩加速濾除雜訊之影像處理裝置、影像處理方法及數位相機
Shang et al. A high dynamic range complementary metal-oxide-semiconductor (CMOS) camera using multi-slope response and an image reconstruction algorithm
JP2015149571A (ja) 画像処理装置、撮像装置及び画像処理プログラム

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: 18902615

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2020522333

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18902615

Country of ref document: EP

Kind code of ref document: A1