A wide-area high-sensitivity detection system based on single-photon time-domain accumulation

Abstract

This invention provides a wide-area high-sensitivity detection system based on single-photon time-domain accumulation. In scanning imaging mode, the intelligent integrated processing unit drives the optical fast scanning component to perform forward scanning, completing the acquisition of ground object information over a large area under the satellite, and then converging the image through the optical lens. The photon counting focal plane receives and forms a photon count matrix representing the ground object information, which is transmitted to the intelligent integrated processing unit. The intelligent integrated processing unit extracts the location information of important targets or areas. Then, the intelligent integrated processing unit drives the optical fast scanning component to perform reverse scanning, similarly converging the image through the optical lens, forming a photon count matrix through the photon counting focal plane, and transmitting it to the intelligent integrated processing unit. Based on the location information of important targets or areas extracted during forward scanning, the intelligent integrated processing unit slices the location of important targets or areas to form image information of important targets or areas.

Description

Technical Field

[0001] This invention belongs to the field of satellite optical remote sensor technology, and relates to a wide-area high-sensitivity detection system based on single-photon time-domain accumulation, which is suitable for spaceborne optical cameras. Background Technology

[0002] With the advancement of space optical remote sensing technology, aerospace applications have placed higher demands on spaceborne optical cameras for high resolution and ultra-wide swaths. This has directly led to increasingly larger camera apertures, fields of view, and overall dimensions, as well as larger focal plane arrays. Taking achieving 0.5m resolution and a 300km swath width at an orbital altitude of 500km as an example: if conventional pushbroom imaging is used, the required camera aperture would typically be at least 0.5m, with a field of view of 33.4°. The required focal plane array would cost at least 600,000 yuan. Even using the smallest known detector pixel size of 3.5μm, the focal plane length would still reach 2.1m, and the system size would exceed 3000mm×3000mm×2500mm. The size, weight, and power consumption far exceed those of conventional cameras, and the resource requirements for the payload far exceed the capabilities of conventional satellites, resulting in an excessively high overall cost. However, if optical... Achieving a swath width of over 300km using machine scanning, while maintaining a camera aperture of over 0.5m, requires a focal plane array size of only about 20,000 yuan, a corresponding field of view of about 1.14°, and a system size of about 2000mm × 750mm × 900mm. Conventional satellite capabilities can meet the resource requirements for this size. However, the extremely short integration time of only 1μs introduces challenges related to high-speed readout and low signal-to-noise ratio. On one hand, the readout line rate needs to reach approximately 1000kHz; on the other hand, based on typical irradiance, the number of photons is only about 10, requiring approximately 1000 integration stages to achieve a 40dB signal-to-noise ratio. Furthermore, current applications for detecting valuable targets in low-light conditions under complex wide-area backgrounds and imaging in dim conditions such as dawn / dusk / low light also demand high sensitivity and a large dynamic range, requiring a dynamic range of 100dB. Based on the above application requirements, the core performance requirements for the system are ultra-high line frequency (1000kHz), ultra-high integration series (1000 series), and ultra-large dynamic range (100dB).

[0003] Currently, traditional detectors—such as TDI CCD and TDI CMOS—are limited by the inherent defects of photoelectric transfer imaging systems: On the one hand, after the received light signal is converted into electrons, the transfer efficiency is generally <1 due to inherent transfer residue when transferring to the next stage of each TDI (time-delay integration). The transfer efficiency will affect the MTF of the same image. In order to reduce the impact of transfer efficiency on the system imaging quality and to improve the detector yield (pixel yield requirement of 100%), the number of integration stages of traditional detectors is generally only 128. On the other hand, the photoelectric transfer process from each TDI stage to the next stage generally requires a transfer time of about 100 ns, which requires encroaching on the exposure time. At the same time, each pixel also requires a certain reading time (including analog-to-digital conversion). Therefore, the line frequency of traditional detectors is generally only 100 kHz. In addition, after the detector performs photoelectric transfer stage by stage, the electrons need to accumulate in the integrating capacitor located below the pixel. However, due to the limitation of the maximum range of the integrating capacitor, there is an inherent upper limit of full-well charge, as well as inherent readout noise. Therefore, the dynamic range of traditional detectors is generally only 80 dB. Therefore, imaging systems based on traditional detectors such as TDI CCD or TDI CMOS are limited by existing systems in terms of line frequency, integration level, and dynamic range, and can no longer meet the application requirements of high resolution, ultra-wide swath, or high sensitivity.

[0004] Furthermore, if the system achieves high resolution and ultra-wide swath, it will generate image data with an extremely high data rate, which will place a huge burden on the existing data transmission capabilities. Taking 0.5m resolution and 300km swath width as an example, with a quantization bit depth of 12 bits, the raw data rate will reach 248.65Gbps. Whether it is transmitted from the camera end to the data transmission end or from the data transmission end to the ground system, it means extremely high resource requirements. Summary of the Invention

[0005] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose a wide-area high-sensitivity detection system based on single-photon time-domain accumulation and optical fast-scan imaging based on time-delay integration. This system breaks through the bottleneck of traditional photoelectric transfer imaging and can achieve high-resolution ultra-wide swath, high-resolution high-sensitivity, and high-resolution video multi-mode imaging in ultra-short integration time at the microsecond or even sub-microsecond level. At the same time, it can also realize the self-search discovery and identification of important targets, which greatly improves the efficiency of optical remote sensors.

[0006] The technical solution of this invention is: a wide-area high-sensitivity detection system based on single-photon time-domain accumulation, comprising an optical fast-scanning assembly, an optical lens, a photon counting focal plane, and an intelligent integrated processing unit; in scanning imaging mode, the intelligent integrated processing unit drives the optical fast-scanning assembly to perform forward scanning, completing the acquisition of ground object information over a large area of ​​the nadir point, and then converging the image through the optical lens. The photon counting focal plane receives and forms a photon count matrix characterizing the ground object information, which is transmitted to the intelligent integrated processing unit. The intelligent integrated processing unit extracts the location information of important targets or areas. Then, the intelligent integrated processing unit drives the optical fast-scanning assembly to perform reverse scanning, similarly converging the image through the optical lens, forming a photon count matrix through the photon counting focal plane, and transmitting it to the intelligent integrated processing unit. Based on the location information of important targets or areas extracted during forward scanning, the intelligent integrated processing unit slices the location of important targets or areas to form image information of important targets or areas.

[0007] The optical fast scanning assembly uses a scanning mirror made of low-density, high-elasticity modulus material; in scanning imaging mode, the scanning angle θ of the optical fast scanning assembly is calculated using the following formula:

[0008]

[0009] In the formula, H is the orbital altitude of the system, R is the Earth's radius, and SW is the system's swath width.

[0010] The optical lens is 3D printed in one piece using a low-density, high-elastic-modulus homogeneous material; the focal length fl of the optical lens is calculated using the following formula:

[0011] IFOV = GSD / H

[0012] fl = pl / IFOV

[0013] In the formula, IFOV is the instantaneous field of view of the system, GSD is the ground pixel resolution of the system's nadir point, and pl is the pixel size of the photon counting focal plane.

[0014] The photon counting focal plane uses a single-photon avalanche photodiode array to count single photons of ground features imaged by the optical lens. The minimum detectable photon count is N. min The dark count is DCR, N min This indicates that the focal plane of the photon counting unit receives N photons per second. min Each photon count increases by 1; DCR represents the random error of the photon counting focal plane in single-photon counting; the single-photon avalanche photodiode array size is K×T, where K represents the number of pixels along the orbital flight linear array direction, and T represents the number of pixels along the cross-track scanning time delay integration direction; in scanning imaging mode, the working line frequency f of the photon counting focal plane is calculated according to the following formula:

[0015]

[0016]

[0017] In the formula, v s Let η be the velocity at the nadir point, G be the gravitational constant, M be the Earth's mass, and η be the scanning efficiency of the optical fast scan module.

[0018] In scanning imaging mode, the intelligent integrated processing unit performs bidirectional scanning drive control on the optical fast scanning component, and the pointing angle difference Δ between the two scans of the optical fast scanning component in the forward and reverse directions... X Δ Y Calculate using the following formula:

[0019]

[0020]

[0021] In scanning imaging mode, the intelligent integrated processing unit processes the photon count matrix acquired by the photon counting focal plane as follows: During forward scanning, the photon count matrix (K×T) is divided into two sets of matrices [K×(1~X)] and {K×[(T-X+1)~T]}. These are then accumulated at X levels to form two sets of matrices (K×1) with a time difference ΔT, thus suitable for target differential detection. During reverse scanning, the photon count matrix (K×T) is accumulated at T levels to form a single matrix (K×1), thereby improving the image signal-to-noise ratio. ΔT is calculated using the following formula:

[0022]

[0023] In the formula, X is a predefined integral series of forward scan time delay.

[0024] In video imaging mode, the intelligent integrated processing unit drives the optical fast scanning component at a speed not less than v s The speed of reverse scanning enables ground video imaging.

[0025] In pushbroom imaging mode, based on the satellite side tilt of 90°, the intelligent integrated processing unit controls the optical fast scanning component to stop scanning, thereby achieving pushbroom imaging of the ground. At this time, the number of pushbroom pixels in the track crossing direction is K, and the time delay integral series along the track direction is T.

[0026] The signal-to-noise ratio (SNR) is calculated using the following formula for different imaging modes:

[0027] In scanning imaging mode,

[0028] In video imaging mode,

[0029] In push-broom imaging mode,

[0030]

[0031] In the formula, N norminal S is the number of photons received per unit time at the focal plane in response to typical illumination, where S is the signal photon count value, N is the noise photon count value, and T0 is the custom integration time in video imaging mode.

[0032] The advantages of this invention compared to the prior art are:

[0033] 1. Based on the characteristics of single-photon avalanche photodiode arrays and their ultra-high-speed response to single photons at the picosecond level, combined with time delay integration, global digital imaging, transfer, and readout are realized.

[0034] 2. For the first time, time-delay integral photon counting imaging was proposed, breaking through the bottleneck of traditional photoelectric transfer imaging system, thereby realizing high-resolution, ultra-wide swath, ultra-high sensitivity and flexible multi-mode video imaging in ultra-short integration time of microseconds or even sub-microseconds.

[0035] 3. For the first time, an optical fast scanning mode based on target identification is proposed. By combining forward and reverse scanning, important targets can be self-searched, discovered, and identified, which greatly improves the efficiency of optical remote sensors. Attached Figure Description

[0036] Figure 1 The working principle of the scanning imaging mode of an optical fast scanning system based on photon counting;

[0037] Figure 2 The working principle of the video imaging mode of an optical fast scanning system based on photon counting;

[0038] Figure 3 The working principle of push-broom imaging mode in an optical fast-scan system based on photon counting;

[0039] Figure 4 This is a configuration diagram of a specific embodiment of the present invention. Detailed Implementation

[0040] like Figure 1-4 As shown, the present invention discloses a wide-area high-sensitivity detection system based on single-photon time-domain accumulation, comprising an optical fast-scan assembly, an optical lens, a photon counting focal plane, and an intelligent integrated processing unit.

[0041] The optical fast scanning component uses a beryllium scanning mirror with low density and high elastic modulus. In scanning imaging mode, the scanning angle of the optical fast scanning component is θ = 43.87° (the orbital altitude of the system is H = 500km, the Earth's radius is R = 6371km, and the swath width is SW = 1000km).

[0042] The optical lens is 3D printed in one piece using low-density, high-elasticity beryllium aluminum alloy material; the focal length of the optical lens is fl = 10m (the system's ground pixel resolution at the nadir point is GSD = 0.5m, the system's instantaneous field of view (IFOV) is 1μrad, and the photon counting focal plane pixel size is pl = 10μm).

[0043] The photon counting focal plane uses a single-photon avalanche photodiode array to count single photons of ground features imaged by the optical lens, with a minimum detectable photon count N. min =2, dark count is DCR=100Hz; single-photon avalanche photodiode array size is K×T, K=25750, T=1000; in scanning imaging mode, the working line frequency of the photon counting focal plane is f=1000kHz (nadir point flight v s =7.0623km / s, optical fast scanning component scanning efficiency η=0.84).

[0044] In scanning imaging mode, the intelligent integrated processing unit performs bidirectional scanning drive control on the optical fast scanning component, and the pointing angle difference Δ between the two scans of the optical fast scanning component in the forward and reverse directions... X =1.2393°, Δ Y =1.4750°; In scanning imaging mode, the intelligent integrated processing unit processes the photon count matrix acquired by the photon counting focal plane as follows: the photon count matrix (25750×1000) in the forward scanning is divided into two sets of matrices [25750×(1~100)] and [25750×(901~1000)], and after accumulating X=100 levels, two sets of matrices (25750×1) with a time difference ΔT=63.72ms are formed, which is suitable for target differential detection. In the reverse scanning, the photon count matrix (25750×1000) is accumulated for 1000 levels to form a set of matrices (25750×1), thereby improving the image signal-to-noise ratio.

[0045] In video imaging mode, the intelligent integrated processing unit drives the optical fast scanning component to perform a reverse scan at a speed of not less than 7.0623 km / s to achieve ground video imaging.

[0046] In pushbroom imaging mode, based on the satellite side tilt of 90°, the intelligent integrated processing unit controls the optical fast scanning component to stop scanning, thereby achieving pushbroom imaging of the ground. At this time, the number of pushbroom pixels in the orbital direction is 25750, and the time delay integral series in the orbital direction is 1000.

[0047] In scanning imaging mode, the system's typical signal-to-noise ratio (SNR) is 100 (N is the typical number of photons N received per unit time at the photon counting focal plane). norminal=20 / μs, at which point the signal photon count S = 10000 and the noise photon count N = 100); in video imaging mode, the typical signal-to-noise ratio SNR = 100 (custom integration time T0 = 1ms, at which point the signal photon count S = 10000 and the noise photon count N = 100); in pushbroom imaging mode, the typical signal-to-noise ratio SNR = 841 (signal photon count S = 708000 and noise photon count N = 100).

[0048] The contents not described in detail in this specification are common knowledge to those skilled in the art.

Claims

1. A wide-area high-sensitivity detection system based on single-photon time-domain accumulation, characterized in that: It includes an optical fast scanning assembly, an optical lens, a photon counting focal plane, and an intelligent integrated processing unit. In scanning imaging mode, the intelligent integrated processing unit drives the optical fast scanning assembly to perform a forward scan, completing the acquisition of ground object information over a large area of ​​the nadir point. The image is then converged by the optical lens, received by the photon counting focal plane, and forms a photon count matrix representing the ground object information. This matrix is ​​then transmitted to the intelligent integrated processing unit, which extracts the location information of important targets or areas. Then, the intelligent integrated processing unit drives the optical fast scanning assembly to perform a reverse scan. Similarly, the image is converged by the optical lens, and a photon count matrix is ​​formed by the photon counting focal plane. This matrix is ​​then transmitted to the intelligent integrated processing unit. Based on the location information of important targets or areas extracted during the forward scan, the intelligent integrated processing unit slices the location of important targets or areas to form image information of the important targets or areas. The photon counting focal plane uses a single-photon avalanche photodiode array to count single photons of ground features imaged by the optical lens. The minimum detectable photon count is N. min The dark count is DCR, N min This indicates that the focal plane of the photon counting unit receives N photons per second. min The number of photons, the photon count value increases by 1, DCR represents the random error of the photon counting focal plane in single photon counting; the size of the single photon avalanche photodiode array is K×T, where K represents the number of pixels in the linear array direction along the track and T represents the number of pixels in the direction of the track scanning time delay integration. In scanning imaging mode, the working line frequency f of the photon counting focal plane is calculated as follows: In the formula, v s η is the velocity at the nadir point, G is the gravitational constant, M is the mass of the Earth, and η is the scanning efficiency of the optical fast scan assembly. The focal length (fl) of an optical lens: IFOV = GSD / H fl = pl / IFOV In the formula, IFOV is the instantaneous field of view of the system, GSD is the ground pixel resolution of the system's nadir point, and pl is the pixel size of the photon counting focal plane. In scanning imaging mode, the intelligent integrated processing unit performs bidirectional scanning drive control on the optical fast scanning component, and the pointing angle difference Δ between the two scans of the optical fast scanning component in the forward and reverse directions... X Δ Y The calculation process includes: In scanning imaging mode, the intelligent integrated processing unit processes the photon count matrix acquired by the photon counting focal plane as follows: during forward scanning, the photon count matrix (K×T) is divided into two sets of matrices [K×(1~X)] and {K×[(T-X+1)~T]}, and after accumulating X levels, two sets of matrices (K×1) with a time difference ΔT are formed, which is suitable for target differential detection. During reverse scanning, the photon count matrix (K×T) is accumulated T levels to form a set of matrices (K×1), thereby improving the image signal-to-noise ratio. The time difference ΔT is calculated using the following formula: In the formula, X is a predefined integral series of forward scan time delay; In video imaging mode, the intelligent integrated processing unit drives the optical fast scanning component at a speed not less than v s The high-speed reverse scanning enables ground video imaging; In pushbroom imaging mode, based on the satellite side tilt of 90°, the intelligent integrated processing unit controls the optical fast scanning component to stop scanning, thereby achieving pushbroom imaging of the ground. At this time, K is the number of pushbroom pixels in the track crossing direction, and T is the time delay integral series along the track direction. The signal-to-noise ratio (SNR) is calculated using the following formula for different imaging modes: In scanning imaging mode, In video imaging mode, In push-broom imaging mode, In the formula, N norminal S is the number of photons received per unit time at the focal plane in response to typical illumination, where S is the signal photon count value, N is the noise photon count value, and T0 is the custom integration time in video imaging mode.

2. The wide-area high-sensitivity detection system based on single-photon time-domain accumulation according to claim 1, characterized in that: The optical fast scanning assembly uses a scanning mirror made of low-density, high-elasticity modulus material; In scanning imaging mode, the scanning angle θ of the optical fast scanning component: In the formula, H is the orbital altitude of the system, R is the Earth's radius, and SW is the system's swath width.

3. The wide-area high-sensitivity detection system based on single-photon time-domain accumulation according to claim 1, characterized in that: The optical lens is 3D printed in one piece using a low-density, high-elasticity homogeneous material.

Images