Three-dimensional stacking technology-based spatio-temporal differential vision sensor chip, and imaging system

Abstract

The present application provides a three-dimensional stacking technology-based spatio-temporal differential vision sensor chip, and an imaging system. The chip comprises a pixel array, a storage circuit, and spatio-temporal differential and quantization circuits. The pixel array comprises a plurality of pixel units, and a photosensitive circuit is provided in the pixel array. By means of three-dimensional stacking, the pixel array is arranged on a top wafer, and the spatio-temporal differential and quantization circuits are arranged on a same bottom wafer, thereby greatly improving the integration level, effectively reducing the area of the chip, and also relieving the output bandwidth pressure. The photosensitive circuit converts acquired optical signals of the pixel units into analog electrical signals of the pixel units; the storage circuit stores the analog electrical signals or digital electrical signals of the pixel units; and the spatio-temporal differential and quantization circuits perform spatio-temporal differential and quantization processing on the analog electrical signals of the pixel units to obtain digital spatio-temporal differential electrical signals of the pixel units. Thus, the spatial and temporal variations of visual signals can be obtained simultaneously, and an efficient and robust visual representation is formed.

Description

Spatiotemporal differential vision sensor chip and imaging system based on three-dimensional stacking technology

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese patent application No. 2023114206704, filed on October 30, 2023, entitled “Spatiotemporal differential visual sensor chip and imaging system based on three-dimensional stacking technology”, which is incorporated herein by reference in its entirety. Technical Field

[0003] The present application relates to the field of visual sensor technology, and in particular to a spatiotemporal differential visual sensor chip and imaging system based on three-dimensional stacking technology. Background Art

[0004] A visual sensor is a photoelectric detection device that converts an optical signal including an image and / or distance (i.e., depth) information of an object into an electrical signal. Various types of visual sensors have been developed, such as APS (Active Pixel Sensor), DVS (Dynamic Vision Sensor), and DAVIS (Dynamic and active pixel vision sensor) to provide high-quality image information of an object. However, on the one hand, all circuit elements of existing visual sensors are manufactured on a two-dimensional wafer, which increases the chip area. On the other hand, existing visual sensors can only output information on the temporal variation of visual signals on the focal plane, which has certain limitations.

[0005] Summary of the Invention

[0006] The present application provides a spatiotemporal differential vision sensor chip and imaging system based on three-dimensional stacking technology, which is used to solve the defects of the existing technology that the two-dimensional vision sensor chip has a large area and only outputs time-varying signals. The present application adopts three-dimensional stacking technology to design a spatiotemporal differential vision sensor chip, which can effectively reduce the chip area and can simultaneously obtain the time and spatial changes of the visual signal, forming an efficient and robust visual representation.

[0007] The present application provides a spatiotemporal differential visual sensor chip based on three-dimensional stacking technology, comprising a pixel array, a storage circuit, and a spatiotemporal differential and quantization circuit; the pixel array is arranged on a top wafer, the spatiotemporal differential and quantization circuit is arranged on the same bottom wafer, and the top wafer and the bottom wafer are connected based on a three-dimensional stacking manner; the pixel array comprises a plurality of pixel units; a photosensitive circuit is arranged in the pixel array; the photosensitive circuit is used to convert the acquired light signal of the pixel unit into an analog electrical signal of the pixel unit; the storage circuit is used to store the analog electrical signal or digital electrical signal of the pixel unit; the spatiotemporal differential and quantization circuit is used to perform spatiotemporal differential and re-quantization processing on the analog electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit, or the spatiotemporal differential and quantization circuit is used to quantize the analog electrical signal of the pixel unit to obtain a digital electrical signal of the pixel unit, and then perform spatiotemporal differential processing on the digital electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit.

[0008] According to a spatiotemporal differential visual sensor chip based on three-dimensional stacking technology provided by the present application, the storage circuit includes multiple storage nodes, and the multiple storage nodes are used to store the analog electrical signals of the pixel units at different positions and different times in a multi-node randomly accessible cache manner; the spatiotemporal differential and quantization circuit includes a time differential and quantization unit and a spatial differential and quantization unit; the time differential and quantization unit is used to perform time differential and quantization processing on the analog electrical signals of the pixel units at different times to obtain the time differential values ​​of the current pixel unit position at different times; the spatial differential and quantization unit is used to perform spatial differential and quantization processing on the analog electrical signals of the pixel units at different positions to obtain the spatial differential values ​​of the pixel at the current pixel unit position and the adjacent pixels at the current time.

[0009] According to a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by the present application, it also includes a pulse signal generator; the pulse signal generator is used to control the exposure of the photosensitive circuit.

[0010] According to a spatiotemporal differential visual sensor chip based on three-dimensional stacking technology provided by the present application, a pulse signal generator is provided in each pixel unit, and the digital spatiotemporal differential electrical signals of each pixel unit are output in the same time interval or adaptively programmable time interval.

[0011] According to a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by the present application, a plurality of pixel units share one pulse signal generator, and the digital spatiotemporal differential electrical signals of each pixel unit are output at the same time interval.

[0012] According to a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by the present application, the spatiotemporal differential and quantization circuit is arranged outside the pixel unit and adopts a column-level signal readout method.

[0013] According to a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by the present application, the spatiotemporal differential and quantization circuit is arranged in each of the pixel units, and a pixel-level signal readout method is adopted.

[0014] According to a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by the present application, the digital electrical signal of the pixel unit is quantized and read out in a multi-valued manner.

[0015] According to a spatiotemporal differential visual sensor chip based on three-dimensional stacking technology provided by the present application, it also includes a fusion circuit, which is used to fuse the analog electrical signals of multiple pixel units to obtain a fused analog electrical signal; the spatiotemporal differential and quantization circuit is also used to perform spatiotemporal differential and quantization processing on the fused analog electrical signal to obtain a fused digital spatiotemporal differential electrical signal.

[0016] The present application also provides an imaging system, comprising the above-mentioned spatiotemporal differential vision sensor chip based on three-dimensional stacking technology.

[0017] The present application provides a spatiotemporal differential visual sensor chip and imaging system based on three-dimensional stacking technology, wherein the chip includes a pixel array, a storage circuit, and a spatiotemporal differential and quantization circuit. The pixel array includes a plurality of pixel units, and a photosensitive circuit is arranged in the pixel array. The pixel array is arranged on the top wafer in a three-dimensional stacking manner, and the spatiotemporal differential and quantization circuit is arranged on the same bottom wafer, which greatly improves the integration level, can effectively reduce the chip area, and alleviates the output bandwidth pressure. The photosensitive circuit converts the acquired pixel unit light signal into an analog electrical signal of the pixel unit; the storage circuit stores the analog electrical signal or digital electrical signal of the pixel unit; the spatiotemporal differential and quantization circuit performs spatiotemporal differential and quantization processing on the analog electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit, which can simultaneously obtain the temporal and spatial variation of the visual signal, forming an efficient and robust visual representation. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] FIG1 is a flow chart of signal processing of a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by an embodiment of the present application;

[0020] FIG2 is a schematic diagram of a synchronous trigger signal provided in an embodiment of the present application;

[0021] FIG3 is a block diagram of a pixel internal structure that meets the spatiotemporal differential output requirements provided by an embodiment of the present application;

[0022] FIG4 is a circuit diagram of a pixel that meets the spatiotemporal differential output requirements provided by an embodiment of the present application;

[0023] FIG5 is a schematic diagram of a pathway architecture of a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided by an embodiment of the present application;

[0024] FIG6 is a diagram of a column-level processing array based on a three-dimensional stacking process provided by an embodiment of the present application;

[0025] FIG7 is a diagram of a pixel-level processing array based on a three-dimensional stacking process provided by an embodiment of the present application;

[0026] FIG8 is an architecture diagram of a high-precision multi-value time-varying visual sensor chip in a full-array asynchronous form provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of this application.

[0028] Currently, most CMOS image sensors (CIS) capture video using a frame-based sampling principle. This means that each CIS image frame records the output of all pixels in the pixel array, and each frame is equally spaced. Because transistors are integrated within the pixels to achieve high-performance charge-to-voltage conversion, CIS is also known as an active pixel sensor (APS). Through the color filter array covering the pixel array, CIS can sense visible light of different wavelengths to produce color images. CIS offers the advantages of high pixel array resolution, excellent color reproduction, and high image quality.

[0029] An event camera, also known as a dynamic visual sensor (DVS), is a new type of imaging system. Unlike traditional cameras that use a shutter to control the frame rate and record light intensity on a per-frame basis for all pixels, an event camera is sensitive to the rate of change of light intensity, and each pixel independently records the change in the logarithm of the light intensity at that pixel. When the change exceeds a threshold, a positive or negative pulse is generated. Due to the asynchronous nature of the event camera, it is not limited by the shutter and has extremely high temporal resolution (a frame rate of approximately 10,000 fps, compared to the approximately 100 fps of traditional cameras). Combined with its sensitivity to change, it is naturally adaptable to tasks such as motion monitoring. Another camera, called DAVIS, combines a traditional activated pixel sensor (APS) with DVS to record both single-frame images and event information, combining the advantages of the high spatial resolution of traditional cameras with the high temporal resolution of DVS cameras.

[0030] On the one hand, in the traditional visual sensor manufacturing process, all circuit elements are manufactured on a two-dimensional wafer. Under this traditional process, the processing circuit in the pixel will occupy a large area. On the other hand, taking DVS as an example, DVS can only output the temporal change information of the visual signal on the focal plane, but this information is very easy to be interfered with. For example, when there is flickering light in the outside world, DVS will fail and will not be able to separate the signal changes caused by light source changes and motion. In addition, the spatial gradient of the image is the basis of many algorithms, such as the Harris corner algorithm, and this information is very important. Although DVS can obtain the spatial gradient of the object's motion to a certain extent, that is, through The gradient can be approximated, but due to its coupling with the object's velocity, it is difficult to obtain accurate results. Therefore, there is an urgent need for visual sensors that can simultaneously perceive temporal and spatial changes.

[0031] Please refer to Figure 1, which is a signal processing flow chart of a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided in an embodiment of the present application.

[0032] In order to solve the technical problems existing in the prior art, the present application provides a spatiotemporal differential visual sensor chip based on three-dimensional stacking technology, including a pixel array, a storage circuit and a spatiotemporal differential and quantization circuit; the pixel array is arranged on a top wafer, and the spatiotemporal differential and quantization circuit is arranged on the same bottom wafer, and the top wafer and the bottom wafer are connected based on a three-dimensional stacking manner; the pixel array includes multiple pixel units; a photosensitive circuit is arranged in the pixel array; the photosensitive circuit is used to convert the acquired pixel unit light signal into an analog electrical signal of the pixel unit; the storage circuit is used to store the analog electrical signal or digital electrical signal of the pixel unit; the spatiotemporal differential and quantization circuit is used to perform spatiotemporal differential and then quantize the analog electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit, or the spatiotemporal differential and quantization circuit is used to quantize the analog electrical signal of the pixel unit to obtain the digital electrical signal of the pixel unit, and then perform spatiotemporal differential processing on the digital electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit.

[0033] The human visual system can perceive temporal changes and spatial changes at the same time, and is more sensitive and robust to the outside world. Inspired by human vision, this application proposes a visual sensor based on temporal changes and spatial gradients. This visual sensing architecture needs to obtain the temporal and spatial changes of the visual signal in a synchronous or asynchronous manner. The current moment signal and the signal at any previous moment, and the photosensitive unit at any position in space, are calculated in the form of differences in the charge, analog or digital domain, and the results are output. The transformation amount is quantized and read out in a high-precision multi-value format (≥1bit). The time difference is referred to as TD, and the spatial difference is referred to as SD. At the same time, based on the three-dimensional stacking process, the photosensitive circuit and the processing circuit (i.e., the differential and quantization circuit) are separated onto different layers of wafers, which greatly improves the photosensitivity, processing performance and energy efficiency of the chip.

[0034] Of course, the visual sensor of the present application may also include a color channel, thereby forming a three-channel bionic visual sensor with the time difference channel and the space difference channel.

[0035] The storage circuit can be set on either the top wafer or the bottom wafer, and this application does not impose any special limitation on this.

[0036] Based on the above embodiment:

[0037] As a preferred embodiment, the storage circuit includes multiple storage nodes, and the multiple storage nodes are used to store the analog electrical signals of pixel units at different positions and at different times in a multi-node randomly accessible cache manner; the spatiotemporal difference and quantization circuit includes a time difference and quantization unit and a space difference and quantization unit; the time difference and quantization unit is used to perform time difference and quantization processing on the analog electrical signals of pixel units at different times to obtain the time difference values ​​of the current pixel unit position at different times; the space difference and quantization unit is used to perform space difference and quantization processing on the analog electrical signals of pixel units at different positions to obtain the space difference value of the pixel at the current pixel unit position and the adjacent pixels at the current moment.

[0038] Specifically, this application provides a multi-channel visual sensor. The two channels are: a temporal differential channel (TD) and a spatial differential channel (SD). Information from the TD and SD channels is independently output from the chip to subsequent image processing modules. By providing both TD and SD channels, this application ensures that the sensor chip has complete temporal and spatial information.

[0039] The TD path outputs the time difference value of the current pixel position (x, y) at different times. The expression of the obtained path output is TD(x, y, t n )=Q TD (I(x,y,t n )-I(x,y,t n-1 ))

[0040] where Q TD It is a high-precision quantization method (≥1 bit) for time difference.

[0041] SD channel outputs the current time t n The spatial difference between the current pixel position (x, y) and the adjacent pixel (diagonal or xy direction). For the xy direction difference, the expression of the obtained path output is SD x (x,y,t n )=Q SD (I(x,y,t n )-I(x+1,y,t n )) SD y (x,y,t n )=Q SD (I(x,y,t n )-I(x,y+1,t n ))

[0042] For oblique differential, the expression of the channel output is obtained as

[0043] where Q SDIt is a high-precision quantization method for spatial difference (≥1 bit).

[0044] All the signals mentioned above are three-dimensional, including the spatial two-dimensional quantities x and y and the time dimension t. n ), SD x (x,y,t n ), SD y (x,y,t n ), The data obtained, I(x,y,t n ) is the absolute size of the corresponding pixel at the current moment.

[0045] Among them, the time t at which the signal is acquired n ,t n-1 ,t n-2 …can be acquired in a manner that the entire array is synchronized with the same time interval, the entire array is synchronized with a variable time interval, or the entire array is asynchronous.

[0046] It should be noted that the multi-node random access cache means that if the analog electrical signal output by the photosensitive circuit is stored in node 1 at this moment, it will be stored in node 2 at the next moment, and in node 4 at the next moment, and so on.

[0047] Please refer to FIG2 , which is a schematic diagram of a synchronous trigger signal provided in an embodiment of the present application.

[0048] As a preferred embodiment, it further includes a pulse signal generator; the pulse signal generator is used to control the exposure of the photosensitive circuit.

[0049] As a preferred embodiment, a pulse signal generator is provided in each pixel unit, and the digital spatiotemporal differential electrical signals of each pixel unit are output in the same time interval or in an adaptively programmable time interval.

[0050] As a preferred embodiment, a plurality of pixel units share a pulse signal generator, and the digital spatiotemporal differential electrical signals of each pixel unit are output at the same time interval.

[0051] Specifically, the present application uses a synchronous pulse generated by a pulse signal generator set outside or inside the pixel unit to trigger the photosensitive circuit to record the visual signal. After receiving the pulse, the photosensitive circuit begins to convert the optical signal into an electrical visual signal, and performs further readout and quantification.

[0052] The synchronization pulse generation method can be set not only to a fixed interval, but also to an adaptive, programmable variable interval. This adaptive interval can adapt to the changing characteristics of the external visual signal. When the amount of change is large and the frequency of change is high, a higher sampling frequency is used, and when the signal is low frequency, a lower sampling frequency is used to reduce data volume and energy consumption.

[0053] This application also supports N×N pixel units forming a macroblock to share an intra-pixel pulse trigger generator to reduce the chip design complexity and occupied area.

[0054] Please refer to Figure 3, which is an internal block diagram of a pixel that meets the spatiotemporal differential output requirements provided by an embodiment of the present application.

[0055] Please refer to FIG4 , which is a circuit diagram of a pixel that meets the spatiotemporal differential output requirements provided by an embodiment of the present application.

[0056] Please refer to FIG5 , which is a path architecture diagram of a spatiotemporal differential vision sensor chip based on three-dimensional stacking technology provided in an embodiment of the present application.

[0057] As a preferred embodiment, the spatiotemporal difference and quantization circuit is arranged outside the pixel unit, and a column-level signal readout method is adopted.

[0058] Specifically, in this embodiment, a column-level readout method is used for the electrical digital signal, that is, the differential and quantization modules are independent of the pixels, and each column shares a differential and quantization module. In order to meet the input of the differential module, two storage nodes need to be designed inside each pixel to output two time signals. This embodiment provides a possible specific circuit design. There are more than one actual circuit schematic diagram. This application mainly emphasizes the logic of the protection block diagram, thereby covering all circuit designs.

[0059] Taking a 3*3 pixel array as an example, the readout method of the time difference and spatial difference signals is:

[0060] SD module ① calculates SD in sequence x (x,y,t n )=Q SD (I(x,y,t n )-I(x+1,y,t n )) SD y (x,y,t n )=Q SD (I(x,y,t n )-I(x,y+1,t n )) SD x (x,y+1,t n )=Q SD (I(x,y+1,t n) - I(x + 1, y + 1, t n )) SD y (x, y + 1, t n ) = Q SD (I(x, y + 1, t n ) - I(x, y + 2, t n ))

[0061] SD module ② calculates SD x (x + 1, y, t n ) = Q SD (I(x + 1, y, t n ) - I(x + 2, y, t n )) SD y (x + 1, y, t n ) = Q SD (I(x + 1, y, t n ) - I(x + 1, y + 1, t n )) SD x (x + 1, y + 1, t n ) = Q SD (I(x + 1, y + 1, t n ) - I(x + 2, y + 1, t n )) SD y (x + 1, y + 1, t n ) = Q SD (I(x + 1, y + 1, t n ) - I(x + 1, y + 2, t n ))

[0062] SD module ③ and so on.

[0063] TD module ④ calculates TD(x, y, t n ) = Q TD [[ID=Z65]](I(x, y, t n ) - I(x, y, t n-1 )) TD(x, y + 1, t n ) = Q TD (I(x, y + 1, t n ) - I(x, y + 1, t n-1 )) TD(x, y + 2, t n ) = Q TD (I(x, y + 2, t n ) - I(x, y + 2, t n-1 ))

[0064] TD module ⑤ calculates TD(x + 1, y, t n ) = Q TD It should be noted that there may be some inaccuracies in the translation due to the complexity and potential ambiguity of the original text, especially in the context of specific technical or domain - specific meanings. If possible, it would be beneficial to have more context or domain knowledge to ensure a more precise translation.(I(x+1,y,t n )-I(x+1,y,t n-1 )) TD(x+1,y+1,t n )=Q TD (I(x+1,y+1,t n )-I(x+1,y+1,t n-1 )) TD(x+1,y+2,t n )=Q TD (I(x+1,y+2,t n )-I(x+1,y+2,t n-1 ))

[0065] The same applies to TD module ⑥.

[0066] Please refer to FIG. 6 , which is a diagram of a column-level processing array based on a three-dimensional stacking process according to an embodiment of the present application.

[0067] In this embodiment, based on the three-dimensional stacking process, multiple pixel units are arranged on the top wafer, and multiple pixel units share the spatiotemporal differential and quantization module. The spatiotemporal differential and quantization module is arranged on the bottom wafer, which can realize column-level signal readout, and also greatly improves the chip's photosensitivity, processing performance and energy efficiency.

[0068] Please refer to FIG. 7 , which is a diagram of a pixel-level processing array based on a three-dimensional stacking process according to an embodiment of the present application.

[0069] As a preferred embodiment, the spatiotemporal difference and quantization circuit is arranged in each pixel unit, and a pixel-level signal readout method is adopted.

[0070] In this embodiment, based on the three-dimensional stacking process, the pixel unit is arranged on the top wafer, and a time-space difference and quantization module is correspondingly set in each pixel unit. The time-space difference and quantization module is arranged on the bottom wafer, which can realize the readout of pixel-level signals, and also greatly improves the photosensitivity, processing performance and energy efficiency of the chip.

[0071] Please refer to Figure 8, which is an architecture diagram of a high-precision multi-value time-varying visual sensor chip in a full-array asynchronous form provided in an embodiment of the present application.

[0072] As a preferred embodiment, the digital electrical signal of the pixel unit is quantized and read out in a multi-valued manner.

[0073] Taking into account that the visual sensor DVS in the prior art is asynchronous information and outputs only timestamp and 1-bit information, it is easily affected by noise interference, has low information volume and low signal-to-noise ratio. Since DVS itself can only output time-varying information and the information is 1 bit, it is easy to encounter extreme situations that do not satisfy the generalized sampling theorem and cannot adapt to complex environments. The present application also implements a high-precision multi-value time-varying visual sensor chip architecture in a full-array asynchronous form. The chip realizes self-triggering (supports internal or external triggering, programmable, adaptive triggering), signal storage, intra-pixel time signal differentiation and quantized readout within a single pixel, and each pixel directly outputs high-precision multi-value time differential and spatial differential (accuracy ≥ 1 bit).

[0074] To support global asynchrony, each pixel in this chip has its own control logic. This control logic, called an intra-pixel pulse trigger generator, adaptively adjusts the triggering timing for calculating the time-difference visual signal based on the light intensity level perceived by the pixel. This unique triggering timing for each pixel achieves global asynchrony.

[0075] As a preferred embodiment, it also includes a fusion circuit, which is used to fuse the analog electrical signals of multiple pixel units to obtain a fused analog electrical signal; the spatiotemporal differential and quantization circuit is also used to perform spatiotemporal differential and quantization processing on the fused analog electrical signal to obtain a fused digital spatiotemporal differential electrical signal.

[0076] This application can also support pixel space fusion to achieve time and space differential perception with larger receptive field, larger spatial scale and higher sensitivity. By sharing the readout switch and storage node, the photoelectricity of multiple pixels is fused (that is, the signal generated by the photodiode is linearly related to the optical signal), and then the time and space differential calculation of the fused photoelectric signal is performed. The fusion method can be summed or averaged. The sum is Average

[0077] The fused pixel range typically includes 2×2, 3×3, etc.

[0078] Then calculate the time-space difference (here I 融合 It can be I ave , or I sum ). TD(x,y,t n )=Q(I 融合 (x,y,t n )-I 融合 (x,y,t n-1 )) SD_X(x,y,t n )=Q(I 融合 (x,y,t n )-I融合 (x-1,y,t n )) SD_Y(x,y,t n )=Q(I 融合 (x,y,t n )-I 融合 (x,y-1,t n ))

[0079] In summary, this application proposes a multi-value, high-precision temporal and spatial variation visual sensor chip architecture, and proposes a variety of signal recording and conversion methods, which can greatly improve the visual sensor's ability to reconstruct temporal and spatial dynamic information with high precision.

[0080] The present application also provides an imaging system, comprising the above-mentioned spatiotemporal differential vision sensor chip based on three-dimensional stacking technology.

[0081] For an introduction to an imaging system provided in this application, please refer to the above chip embodiment, and this application will not go into details here.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A spatiotemporal differential visual sensor chip based on three-dimensional stacking technology, including a pixel array, a storage circuit, and a spatiotemporal differential and quantization circuit; The pixel array is arranged on a top wafer, the spatiotemporal difference and quantization circuit is arranged on the same bottom wafer, and the top wafer and the bottom wafer are connected based on a three-dimensional stacking manner; The pixel array includes a plurality of pixel units; a photosensitive circuit is arranged in the pixel array; The photosensitive circuit is used to convert the acquired light signal of the pixel unit into an analog electrical signal of the pixel unit; The storage circuit is used to store the analog electrical signal or digital electrical signal of the pixel unit; The spatiotemporal differential and quantization circuit is used to perform spatiotemporal differential and re-quantization processing on the analog electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit, or the spatiotemporal differential and quantization circuit is used to quantize the analog electrical signal of the pixel unit to obtain a digital electrical signal of the pixel unit, and then perform spatiotemporal differential processing on the digital electrical signal of the pixel unit to obtain a digital spatiotemporal differential electrical signal of the pixel unit.

2. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 1, wherein: The storage circuit includes a plurality of storage nodes, and the plurality of storage nodes are used to store the analog electrical signals of the pixel units at different positions and at different times in a multi-node random access cache manner; The time-space difference and quantization circuit includes a time difference and quantization unit and a space difference and quantization unit; The time difference and quantization unit is used to perform time difference and quantization processing on the analog electrical signals of the pixel unit at different times to obtain time difference values ​​at different times of the current pixel unit position; The spatial difference and quantization unit is used to perform spatial difference and quantization processing on the analog electrical signals of the pixel units at different positions to obtain the spatial difference value between the pixel at the current pixel unit position and the adjacent pixels at the current moment.

3. According to claim 1, the spatiotemporal differential visual sensor chip based on three-dimensional stacking technology further includes a pulse signal generator; the pulse signal generator is used to control the exposure of the photosensitive circuit.

4. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 3, wherein: A pulse signal generator is provided in each pixel unit, and the digital spatiotemporal differential electrical signals of each pixel unit are output in the same time interval or in an adaptive programmable time interval.

5. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 3, wherein: A plurality of the pixel units share one pulse signal generator, and the digital spatiotemporal differential electrical signals of the pixel units are output at the same time interval.

6. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 1, wherein: The spatiotemporal difference and quantization circuit is arranged outside the pixel unit and adopts a column-level signal readout method.

7. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 1, wherein: The spatiotemporal difference and quantization circuit is arranged in each of the pixel units, and adopts a pixel-level signal readout method.

8. The spatiotemporal differential vision sensor chip based on three-dimensional stacking technology according to claim 1, wherein: The digital electrical signal of the pixel unit is quantized and read out in a multi-valued manner.

9. The spatiotemporal differential visual sensor chip based on three-dimensional stacking technology according to any one of claims 1 to 8, further comprising a fusion circuit, wherein the fusion circuit is used to fuse the analog electrical signals of the plurality of pixel units to obtain a fused analog electrical signal; The time-space difference and quantization circuit is also used to perform time-space difference and quantization processing on the fused analog electrical signal to obtain a fused digital time-space difference electrical signal.

10. An imaging system, comprising the spatiotemporal differential vision sensor chip based on three-dimensional stacking technology as described in any one of claims 1 to 9.

Images