A novel intelligent perception system for aerospace moving targets under complex background conditions

Abstract

The present invention provides an intelligent perception system for aerospace moving targets under a new type of complex background condition, comprising: a high-sensitivity optical lens and a pixel cluster intelligent detector, wherein: the pixel cluster intelligent detector includes an information generation module and an information intelligent processing and feedback module; the information generation module performs photoelectric conversion and integration processing on the collected optical signals to generate a number of initial pixel signals; generates a working mode signal and a number of pixel information; the information intelligent processing and feedback module processes the received number of initial pixel signals to generate a feedback signal; judges and processes the number of pixel information according to the working mode signal, and outputs full-frame information or in-window moving target information. The present invention not only has the detection ability of a large dynamic range, but also can effectively capture, detect and track dim aerospace targets under complex backgrounds.

Description

Technical Field

[0001] The present invention belongs to the technical field of aerospace optical remote sensing, and relates to an intelligent perception system for aerospace moving targets under new complex background conditions. Background Art

[0002] The optical characteristics, motion characteristics, etc. of different types of aerospace targets in different flight stages are very different. At present, the space-based infrared optical systems for detecting aerospace moving targets at home and abroad mainly adopt the link mode of target-detection-optical-electric conversion-on-orbit storage-space-ground transmission-ground processing-application. In order to achieve wide coverage, traditional space-based infrared optical systems generally adopt large-scale detector arrays, and the overall readout and transmission are carried out with unified parameters within the full frame. The optical-electric conversion and readout are carried out in sequence, and the invalid background information in the field of view needs to be stored and transmitted at the same time, and on-chip real-time processing cannot be realized. The data volume is large, the information transmission link is long, and effective and fast transmission and distribution cannot be achieved. Detecting aerospace moving targets belongs to target detection under complex backgrounds. The characteristics of different targets vary greatly, and at the same time, the distribution of complex backgrounds also makes the dynamic range of the same scene very large. It is very difficult for traditional detection systems to take into account both low-end detection and high-end detection at the same time, and it is very difficult to achieve detection with a large dynamic range under complex backgrounds, and it is also very difficult to take into account multiple types of targets and multiple models of targets. Space-based aerospace moving target detection is for ultra-long-distance point target detection, especially for dim targets. The detectable energy is limited and is greatly affected by the background, detection system, and transmission link. The detector focal plane of traditional space-based infrared detection systems generally consists of two major parts: a detector chip and a readout circuit. The readout circuit generally adopts the "analog integration" technology, with poor anti-interference ability, and various factors of instability will introduce noise, resulting in limited detection ability.

[0003] Therefore, it is urgent to carry out research on the technology of an intelligent perception system for aerospace moving targets under new complex background conditions, further improve the detection and tracking ability of space-based infrared aerospace moving targets in China, and lay a foundation for the effective and fast transmission and distribution of space-based infrared information. Summary of the Invention

[0004] The technical problem solved by the present invention: Overcoming the deficiencies of the prior art, a novel intelligent perception system for aerospace moving targets under complex background conditions is proposed, which is for detecting and perceiving multiple types and models of aerospace targets, changing the design concept of traditional space-based infrared optical payloads for detecting "static" scenes, carrying out research on an intelligent space-based infrared aerospace target detection-inspection-tracking integrated system, and further strengthening the capabilities of existing and planned systems.

[0005] The technical solution of the present invention:

[0006] An intelligent perception system for aerospace moving targets under new complex background conditions, including a high-sensitivity optical lens and a pixel cluster intelligent detector, wherein:

[0007] The high-sensitivity optical lens converges the optical signal onto the pixel cluster intelligent detector;

[0008] The pixel cluster intelligent detector includes an information generation module and an information intelligent processing and feedback module;

[0009] The information generation module performs photoelectric conversion and integration processing on the collected optical signal, generates a number of initial pixel signals, and sends them to the information intelligent processing and feedback module; according to external instructions and the feedback signal sent by the information intelligent processing and feedback module, generates a working mode signal and a number of pixel information, and sends them to the information intelligent processing and feedback module;

[0010] The information intelligent processing and feedback module processes the received number of initial pixel signals to generate a feedback signal and sends it to the information generation module; receives the working mode signal and a number of pixel information sent by the information generation module, judges and processes the number of pixel information according to the working mode signal, and outputs full-frame information or in-window moving target information.

[0011] In the above sensing system, the information generation module includes a plurality of pixels, and each pixel includes a photoelectric conversion unit, an analog-to-digital conversion unit, a counter, and a controller; where:

[0012] The photoelectric conversion unit performs photoelectric conversion on the collected optical information, generates an analog signal, and sends it to the analog-to-digital conversion unit;

[0013] The analog-to-digital conversion unit converts the analog signal into a digital signal according to the control signal transmitted by the controller and sends it to the counter;

[0014] The controller generates a control signal according to external instructions and the feedback signal sent by the information intelligent processing and feedback module, and sends it to the analog-to-digital conversion unit and the counter;

[0015] The counter counts the digital signal sent by the analog-to-digital conversion unit according to the control signal sent by the controller, generates pixel information, and sends it to the information intelligent processing and feedback module.

[0016] In the above sensing system, the generating of the working mode signal and a number of pixel information according to external instructions and the feedback signal sent by the information intelligent processing and feedback module is specifically:

[0017] According to external instructions, select the target and background characteristic acquisition mode or the feature target detection mode;

[0018] When the working mode signal is the target and background characteristic acquisition mode, according to the feedback signal, perform parameter adaptive adjustment, and through the pixel-level unidirectional integration method, accumulate and count the voltage signal after photoelectric conversion to achieve the cumulative acquisition of energy. After the acquisition is completed, output a number of pixel information of the full-frame information;

[0019] When the working mode signal is the feature target detection mode, according to the feedback signal, slicing windowing and parameter adaptive adjustment are performed. The pixel-level bidirectional integration method is adopted to output several pixel information of the moving target information within the window, realizing the on-chip detection of the moving target.

[0020] In the above sensing system, according to the feedback signal, slicing windowing and parameter adaptive adjustment are performed. The pixel-level bidirectional integration method is adopted to output several pixel information of the moving target information within the window, specifically as follows:

[0021] A pixel cluster composed of m×n pixels is formed with the pixel where the target is located as the center; where m and n are integers greater than or equal to 1;

[0022] Through intelligent sliding windowing, the pixel cluster moves with the movement of the target, and the target information is output with the pixel cluster as the basic unit;

[0023] According to the feedback signal, the reference voltage, forward integration period, reverse integration period, and integration time are set;

[0024] According to the reference voltage, forward integration period, reverse integration period, and integration time, the pixel-level bidirectional integration method is adopted to filter out the background information and output several pixel information of the moving target information within the window.

[0025] In the above sensing system, the pixel-level unidirectional integration method is specifically as follows:

[0026] Step S31: Start the pixel-level unidirectional integration mode;

[0027] Step S32: Set the reference voltage and integration time by the control signal generated by the information intelligent processing and feedback module;

[0028] Step S33: The charges generated by photoelectric conversion are cumulatively integrated. When the integration voltage reaches the designed reference voltage, 1 cumulative count is performed;

[0029] Step S34: Determine whether the integration time has ended. If not, repeat step S33. If so, output the total number of integration times.

[0030] In the above sensing system, the pixel-level bidirectional integration method is specifically as follows:

[0031] Step S41: Start the pixel-level bidirectional integration mode;

[0032] Step S42: Set the reference voltage, forward integration period, reverse integration period, and integration time; the conditions satisfied by the integration time are: Where, t int is the integration time, v is the target movement speed, and GSD is the pixel spatial resolution;

[0033] Step S43: The charges generated by photoelectric conversion are cumulatively integrated. When the integrated voltage reaches the designed reference voltage, a forward count is performed once.

[0034] Step S44: Determine whether the forward integration period has ended. If not, repeat Step S43. If so, proceed to Step S45.

[0035] Step S45: Perform reverse integration. The charges generated by photoelectric conversion are cumulatively integrated. When the integrated voltage reaches the designed reference voltage, a reverse count is performed once.

[0036] Step S46: Determine whether the reverse integration period has ended. If not, repeat Step S45. If so, proceed to Step S47.

[0037] Step S47: Determine whether the integration time has ended. If not, repeat Steps S43 - S46. If so, record the difference between the forward count and the reverse count.

[0038] In the above - mentioned sensing system, judging and processing several pixel information according to the working mode signal and outputting full - frame information or in - window moving target information specifically includes:

[0039] When the working mode signal is the target - and - background characteristic acquisition mode, judge whether several pixel information meet the output requirements. If so, perform data arrangement on several pixel information and output full - frame information. If not, adjust the output feedback signal.

[0040] When the working mode signal is the characteristic target detection mode, judge whether several pixel information meet the output requirements. If so, perform data arrangement on several pixel information and output in - window moving target information. If not, adjust the output feedback signal.

[0041] In the above - mentioned sensing system, the information intelligent processing and feedback module processes the received digital signal to generate a feedback signal, specifically:

[0042] Perform background analysis, target moving position judgment, target moving state analysis, and target optical characteristic analysis on several pixel information input under different working modes to generate a feedback signal. In the target - and - background characteristic acquisition mode, the feedback signal includes the reference voltage and the integration time. In the characteristic target detection mode, the feedback signal includes the window - opening range, the reference voltage, the forward integration period, the reverse integration period, and the integration time.

[0043] In the above - mentioned sensing system, the conditions satisfied by the integration time are:

[0044]

[0045] where tint where \(t\) is the integration time, \(v\) is the target motion speed, and GSD is the pixel spatial resolution.

[0046] In the above perception system, the data arrangement of several pixel information is specifically as follows: the pixel information includes the spatial position of the information pixel generated by the pixel and the time characteristics of pixel sampling. According to the spatial position and sampling time characteristics of each pixel, the pixel information is arranged in a conventional format to output full-frame information or information of moving targets within the window.

[0047] The advantages of the present invention compared with the prior art are as follows:

[0048] (1) The present invention can realize the adaptive adjustment of detection parameters and real-time on-chip information processing. This system integrates two major modules of information generation, intelligent information processing and feedback on-chip, can perform information processing on-chip, and adaptively adjust detection parameters according to detection targets and backgrounds, so that the system is always in the optimal detection efficiency for dynamic targets. At the same time, the unique pixel-level bidirectional integration function of the present invention can couple the spatio-temporal domain information of moving targets in a single-frame image and effectively suppress the influence of complex backgrounds, which is beneficial to quickly detecting targets on-chip and realizing the integration of detection and inspection.

[0049] (2) The present invention can realize the simultaneous detection and tracking of multiple targets. Through the intelligent sliding window function, multiple "pixel cluster" control units can be planned and formed on the detector according to the number of targets, and the parameters within the cluster are automatically set and adjusted according to the target characteristics detected and tracked by it; and due to the pixel-level bidirectional integration function, the targets within the cluster are automatically detected, greatly reducing the information processing pressure brought by the simultaneous detection and tracking of multiple targets.

[0050] (3) The present invention has the ability to detect targets with a large dynamic range and detect weak targets. The present invention has both pixel-level unidirectional integration function and pixel-level bidirectional integration function at the same time. The new detector processing circuit performs photon counting on the signal, and the readout circuit directly outputs digital signals. Pixel-level integration can achieve a charge processing capacity in the gigabit range, can detect extremely "strong" targets, and at the same time, through the accumulation of signals, the energy of weak targets can be accumulated to improve the detection sensitivity of weak targets; in addition, pixel-level bidirectional integration can also extend the detection of "point" targets to the detection of "line" targets, further improving the detection ability of weak targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the system composition and working mode of the present invention

[0052] Figure 2 Schematic diagram of the composition of the pixel cluster intelligent detector of the present invention

[0053] Figure 3 Schematic diagram of the optical system of the present invention

[0054] Figure 4 Schematic diagram of the pixel cluster windowing mode of the present invention

[0055] Figure 5 Principle diagram of pixel-level bidirectional integration of the present invention

[0056] Figure 6 Schematic diagram of the pixel-level bidirectional integration timing of the present invention Detailed implementation manners

[0057] The working principle and process of the present invention will be further explained and described below with reference to the accompanying drawings.

[0058] The present invention discloses an intelligent space and air moving target perception system under a new type of complex background condition, including a high-sensitivity optical lens and a pixel cluster intelligent detector, wherein:

[0059] The high-sensitivity optical lens converges the optical signal onto the pixel cluster intelligent detector;

[0060] The pixel cluster intelligent detector includes an information generation module and an information intelligent processing and feedback module;

[0061] The information generation module performs photoelectric conversion and integration processing on the collected optical signal to generate a number of initial pixel signals, and sends them to the information intelligent processing and feedback module; according to external instructions and feedback signals sent by the information intelligent processing and feedback module, generates a working mode signal and a number of pixel information, and sends them to the information intelligent processing and feedback module;

[0062] The information intelligent processing and feedback module processes the received number of initial pixel signals to generate a feedback signal, and sends it to the information generation module; receives the working mode signal and a number of pixel information sent by the information generation module, judges and processes the number of pixel information according to the working mode signal, and outputs full-frame information or in-window moving target information.

[0063] The information intelligent processing and feedback module processes the received digital signal to generate a feedback signal, specifically:

[0064] Performs background analysis, target movement position judgment, target movement state analysis, and target optical characteristic analysis on the number of pixel information input under different working modes to generate a feedback signal; in the target and background characteristic acquisition mode, the feedback signal includes a reference voltage and an integration time; in the characteristic target detection mode, the feedback signal includes a windowing range, a reference voltage, a forward integration period, a reverse integration period, and an integration time.

[0065] The information generation module includes a plurality of pixels, and each pixel includes a photoelectric conversion unit, an analog-to-digital conversion unit, a counter, and a controller; wherein:

[0066] The optoelectronic conversion unit performs optoelectronic conversion on the collected optical information to generate an analog signal and sends it to the analog-to-digital conversion unit;

[0067] The analog-to-digital conversion unit converts the analog signal into a digital signal according to the control signal transmitted by the controller and sends it to the counter;

[0068] The controller generates a control signal according to the external instruction and the feedback signal sent by the information intelligent processing and feedback module, and sends it to the analog-to-digital conversion unit and the counter;

[0069] The counter counts the digital signal sent by the analog-to-digital conversion unit according to the control signal sent by the controller, generates pixel information, and sends it to the information intelligent processing and feedback module.

[0070] According to the external instruction and the feedback signal sent by the information intelligent processing and feedback module, generate a working mode signal and several pixel information, specifically:

[0071] According to the external instruction, select the target and background characteristic acquisition mode or the feature target detection mode;

[0072] When the working mode signal is the target and background characteristic acquisition mode, according to the feedback signal, perform parameter adaptive adjustment, and accumulate and count the voltage signal after optoelectronic conversion through the pixel-level unidirectional integration method to achieve the cumulative acquisition of energy. After the acquisition is completed, output several pixel information of the full-frame information;

[0073] When the working mode signal is the feature target detection mode, according to the feedback signal, perform slicing windowing and parameter adaptive adjustment, and adopt the pixel-level bidirectional integration method to output several pixel information of the moving target information within the window, realizing the on-chip detection of the moving target.

[0074] According to the feedback signal, perform slicing windowing and parameter adaptive adjustment, and adopt the pixel-level bidirectional integration method to output several pixel information of the moving target information within the window, specifically:

[0075] Form a pixel cluster composed of m×n pixels with the pixel where the target is located as the center; where m and n are integers greater than or equal to 1; through intelligent sliding windowing, make the pixel cluster move with the movement of the target, and output the target information with the pixel cluster as the basic unit; according to the feedback signal, set the reference voltage, forward integration period, reverse integration period, and integration time; according to the reference voltage, forward integration period, reverse integration period, and integration time, adopt the pixel-level bidirectional integration method to filter out the background information and output several pixel information of the moving target information within the window.

[0076] The pixel-level unidirectional integration method is specifically:

[0077] Step S31: Start the pixel-level unidirectional integration mode;

[0078] Step S32: Set the reference voltage and integration time with the control signal generated by the information intelligent processing and feedback module;

[0079] Step S33: Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 addition count;

[0080] Step S34: Determine whether the integration time has ended. If not, repeat Step S33. If so, output the total number of integration times.

[0081] Adopt the pixel-level bidirectional integration method, specifically:

[0082] Step S41: Start the pixel-level bidirectional integration mode;

[0083] Step S42: Set the reference voltage, forward integration period, reverse integration period and integration time; The conditions satisfied by the integration time are: where t int is the integration time, v is the target motion speed, and GSD is the pixel spatial resolution;

[0084] Step S43: Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 forward count;

[0085] Step S44: Determine whether the forward integration period has ended. If not, repeat Step S43. If so, enter Step S45;

[0086] Step S45: Perform reverse integration. Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 reverse count;

[0087] Step S46: Determine whether the reverse integration period has ended. If not, repeat Step S45. If so, enter Step S47;

[0088] Step S47: Determine whether the integration time has ended. If not, repeat Steps S43 to S46. If so, record the difference between the forward count and the reverse count.

[0089] Judge and process the information of several pixels according to the working mode signal, and output the full-frame information or the information of the moving target within the window, specifically:

[0090] When the working mode signal is the target-background characteristic acquisition mode, judge whether the information of several pixels meets the output requirements. If so, perform data arrangement on the information of several pixels and output the full-frame information; If not, adjust the output feedback signal;

[0091] When the working mode signal is the feature target detection mode, determine whether a number of pixel information meets the output requirements. If so, perform data arrangement on the number of pixel information and output the moving target information within the window; if not, adjust the output feedback signal.

[0092] Perform data arrangement on a number of pixel information. Specifically, the pixel information includes the information pixel spatial position generated by the pixel and the time characteristics of pixel sampling. According to the spatial position and sampling time characteristics of each pixel, arrange the pixel information in a predefined format to output the full-frame information or the moving target information within the window.

[0093] Embodiment

[0094] As Figure 1 As shown, in this embodiment, a space-based airborne moving target intelligent perception system under a new type of complex background condition mainly includes: a high-sensitivity optical lens and a pixel cluster intelligent detector, etc. The refrigeration module, temperature control module, power supply and distribution template, etc. are common components of the high-sensitivity infrared detection system and will not be specifically described here. The working spectral band is in the medium and short wave spectral bands and can be extended according to the target and background characteristics.

[0095] As Figure 2 As shown, the pixel cluster intelligent detector includes an information generation module and an information intelligent processing and feedback module. The information generation module consists of multiple pixels, and each pixel consists of a photoelectric conversion unit, an analog-to-digital conversion unit, a counter, a controller, etc. The photoelectric conversion unit performs photoelectric conversion on the collected optical information. The analog-to-digital conversion unit converts the analog signal into a digital signal. The controller receives external instructions and the feedback signal sent by the information intelligent processing and feedback module, and sends control signals to the analog-to-digital conversion unit and the counter. The counter counts the signals sent by the analog-to-digital conversion unit according to the control signal sent by the controller.

[0096] The information intelligent processing and feedback module processes the received signals to generate feedback signals and sends them to the controllers of each pixel in the information generation module; judges and processes the full-frame information or the moving target information within the window according to different working modes.

[0097] The high-sensitivity optical lens receives the target / background infrared signal and converges it onto the detector.

[0098] As Figure 3 As shown, the high-sensitivity optical lens includes a mirror group, a transmissive lens group, and a filter window. Among them, the mirror group includes mirror L1 and mirror L2, the transmissive lens group includes lenses L3, L4, L5, L6, and the window filter L7 is a medium and short wave spectral band pass filter fixed at the front end of the detector packaging structure.

[0099] The materials of lens L3, lens L4, and lens L6 are silicon lenses, and the material of lens L6 is a germanium lens; each lens is separately designed for thermal elimination.

[0100] The incident light passes through mirror L1, mirror L2, lens L3, lens L4, lens L5, and lens L6 in sequence, and then enters the detector through window filter L7.

[0101] Table 1 Structural parameters of high-sensitivity optical lens

[0102]

[0103] The average optical transmittance of the lens module within the working spectral band is 0.72, and the energy concentration is 0.8.

[0104] The system has a target and background characteristic acquisition mode and a characteristic target detection mode. The pixel cluster intelligent detector outputs full-frame information or target information according to the working mode.

[0105] When the working mode is the target and background characteristic acquisition mode, the pixel-level unidirectional integration method is adopted. The optical information transmitted through the optical lens is first collected according to the initial parameters on the information generation module of the pixel cluster intelligent detector, and the generated initial information is transmitted to the information intelligent processing and feedback module for information processing. The information processing system analyzes the energy distribution characteristics, motion characteristics, etc. of the target and the background according to the received information, and generates control signals such as reference voltage and integration time in combination with the evaluation of the information quality, and feeds them back to the information generation module for parameter adjustment to obtain better-quality target and background characteristic data. When the newly collected information meets the output requirements, the full-frame information is output after data arrangement; when it does not meet the output requirements, the newly collected information will be transmitted to the information processing system again for processing and analysis to generate optimized control signals, so that the information generation module performs data collection again with new parameter settings until the requirements are met. The information collected in the target and background characteristic acquisition mode is mainly used to support the construction of the target and background database.

[0106] In this embodiment, the charges generated by photoelectric conversion are cumulatively integrated. When the integration voltage reaches the designed reference voltage, 1 accumulation count is performed, and then the next integration starts. When the integration voltage reaches the reference voltage again, another accumulation count is performed; the above process is repeated continuously during a long integration time. During several repetitions of integration and counting, there is no need to store the signal charges each time, and only the total number of integrations needs to be recorded. This can achieve large dynamic range detection. During this process, if there are changes in moving targets or backgrounds, the information intelligent processing and feedback module will timely adjust and optimize the control signals according to the changes in the detection scene. Finally, after the information intelligent processing and feedback module determines that the detected information meets the output requirements, the full-frame information is output.

[0107] When the working mode is the feature target detection mode, the pixel-level bidirectional integration method is adopted, and an intelligent sliding window system is used. The optical information transmitted through the optical lens is first collected according to the initial parameters on the information generation module of the pixel cluster intelligent detector, and the generated initial information is transmitted to the information intelligent processing and feedback module for information processing. When a moving target appears in the field of view, the information processing system will generate a windowing range control signal according to the position where the target appears and feedback it to the information generation module. One or more m×n "pixel clusters" centered on the target are planned and formed on the information generation module. The information processing system will also analyze the background suppression effect, target optical characteristics, motion characteristics, etc. based on the received information data, and generate a reference voltage, forward integration period, reverse integration period, and integration time control signal in combination with the evaluation of the information quality, and feedback it to the pixels within the pixel cluster of the information generation module for parameter adjustment to obtain a better background suppression effect and higher-quality target information. When the moving target information within the newly collected pixel cluster meets the output requirements, the moving target information is output after data arrangement; when it does not meet the output requirements, the newly collected information will be transmitted to the information processing system again for further processing and analysis to generate optimized control signals, so that the information generation module can be reset with new parameters to continue data collection until the requirements are met. The feature target detection mode can filter out background information, and the effective information it generates is a small-range data corresponding to the windowing range, which can greatly reduce the data volume. At the same time, the information intelligent processing and feedback module can generate multiple windowing range control signals and in-window pixel parameter control signals according to the number of targets, so that the information generation module generates multiple pixel clusters, and the parameters of the pixels within each pixel cluster are different according to the targets they detect. The information collected in the feature target detection mode is mainly used for the actual task execution of feature target detection, tracking, and monitoring.

[0108] In this embodiment, when the system is in this working mode, the system enables the pixel-level integration function. After detecting the target, the information intelligent processing and feedback module determines the target position by processing the information. As Figure 4 shown, according to the determined target position, a window is opened in a 100×100 (m×n) pixel range centered on the target, and the parameters within the window are adaptively set for parameters such as reference voltage, forward and reverse integration periods, and integration time according to the preliminary results of processing information such as target motion and optical characteristics. During the working process, the information intelligent processing and feedback module will dynamically adjust and optimize these parameters according to the in-window moving target information received.

[0109] Adopting the pixel-level unidirectional integration method, it breaks through the traditional way of integrating and quantifying analog voltages by detectors, converts photocurrent into digital signals for integration and quantification, enabling the detector to achieve a gigabit-level charge handling capacity.

[0110] Adopt the pixel-level bidirectional integration method. Break through the traditional way of integrating and quantifying analog voltage by the detector, convert the photocurrent into a digital signal for bidirectional integration and quantification. By controlling the reference voltage, forward and reverse integration periods, integration time, etc., the spatial position and time of the pixel points with energy changes during the detection process can be directly recorded, and automatic filtering of the static background and automatic detection of moving targets can be realized on the chip.

[0111] As Figure 5 and Figure 6 shown, when the system works with the pixel-level bidirectional integration function, the counter records the incremental pulses. The charges generated by photoelectric conversion are cumulatively integrated. When the integration voltage reaches the designed reference voltage, a forward count is performed once, and this process is repeated until the forward integration period ends; then, the charges continue to be cumulatively integrated. When the integration voltage reaches the designed reference voltage, a reverse count is performed once, and this process is repeated until the reverse integration period ends; the above process is repeated until the entire integration time ends. By recording the difference between the forward and reverse integrations within the integration time, the static information in the information can be filtered out, and the information of the dynamic target is output. Repeating the above steps can achieve the detection of moving targets. In this way, the transmission pressure of invalid information can be effectively reduced, and at the same time, the target detection probability and tracking accuracy can be improved.

[0112] This embodiment can be applied to the space-based detection and tracking of various types of aerospace moving targets, solving the problems in China such as the inability to detect multiple types of targets simultaneously, the insensitivity of space-based detection of aerospace dim moving targets, unstable tracking, a large amount of invalid information, and the difficulty of effectively, quickly transmitting and distributing information. At the same time, through the pixel-level bidirectional integration system proposed in this embodiment, the on-board information automatic processing ability will also be greatly improved.

[0113] Although the present invention has been disclosed above with preferred embodiments, it is not used to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modification, equivalent change, and modification made to the above embodiments according to the technical essence of the present invention without departing from the technical solution of the present invention all belong to the protection scope of the technical solution of the present invention.

[0114] The content not detailedly described in the specification of the present invention belongs to the well-known technology of those skilled in the art.

Claims

1. An intelligent perception system for aerospace moving targets under a new type of complex background conditions, characterized in that: It includes a high-sensitivity optical lens and a pixel cluster intelligent detector, where: The high-sensitivity optical lens converges optical signals onto the pixel cluster intelligent detector; The pixel cluster intelligent detector includes an information generation module and an information intelligent processing and feedback module; The information generation module performs photoelectric conversion and integration processing on the collected optical signals to generate a number of initial pixel signals and sends them to the information intelligent processing and feedback module; according to external instructions and feedback signals sent by the information intelligent processing and feedback module, it generates a working mode signal and a number of pixel information and sends them to the information intelligent processing and feedback module; The information intelligent processing and feedback module processes the received number of initial pixel signals to generate feedback signals and sends them to the information generation module; receives the working mode signal and a number of pixel information sent by the information generation module, judges and processes the number of pixel information according to the working mode signal, and outputs full-frame information or in-window moving target information; The information intelligent processing and feedback module processes the received number of initial pixel signals to generate feedback signals, specifically: For the number of pixel information input under different working modes, it performs background analysis, target movement position judgment, target movement state analysis and target optical characteristic analysis to generate feedback signals; in the target and background characteristic acquisition mode, the feedback signals include reference voltage and integration time; in the characteristic target detection mode, the feedback signals include windowing range, reference voltage, forward integration period, reverse integration period and integration time.

2. The intelligent air and space moving target perception system under a new complex background condition according to claim 1, wherein: The information generation module includes multiple pixels, and each pixel includes a photoelectric conversion unit, an analog-to-digital conversion unit, a counter and a controller; where: The photoelectric conversion unit performs photoelectric conversion on the collected optical information to generate an analog signal and sends it to the analog-to-digital conversion unit; The analog-to-digital conversion unit converts the analog signal into a digital signal according to the control signal transmitted by the controller and sends it to the counter; The controller generates a control signal according to external instructions and feedback signals sent by the information intelligent processing and feedback module and sends it to the analog-to-digital conversion unit and the counter; The counter counts the digital signal sent by the analog-to-digital conversion unit according to the control signal sent by the controller to generate pixel information and sends it to the information intelligent processing and feedback module.

3. The intelligent perception system for aerospace moving targets under a new type of complex background conditions according to claim 1, characterized in that: The generation of the working mode signal and a number of pixel information according to external instructions and feedback signals sent by the information intelligent processing and feedback module, specifically: According to external instructions, select the target and background characteristic acquisition mode or the characteristic target detection mode; When the working mode signal is the target and background characteristic acquisition mode, according to the feedback signal, perform parameter adaptive adjustment, and through the pixel-level one-way integration method, accumulate and count the voltage signal after photoelectric conversion to achieve the cumulative acquisition of energy, and output the pixel information of the full-frame information after the acquisition is completed; When the working mode signal is the characteristic target detection mode, according to the feedback signal, perform slicing windowing and parameter adaptive adjustment, adopt the pixel-level two-way integration method, output the pixel information of the in-window moving target information, and achieve the on-chip detection of the moving target.

4. A space-air moving target intelligent perception system under a new type of complex background condition according to claim 3, characterized in that: Based on the feedback signal, perform slice windowing and parameter adaptive adjustment, adopt the pixel-level bidirectional integration method, and output several pixel information of the moving target information within the window. Specifically: Form a pixel cluster composed of m×n pixels with the pixel where the target is located as the center; where m and n are integers greater than or equal to 1; Through intelligent sliding windowing, make the pixel cluster move with the movement of the target, and output the target information with the pixel cluster as the basic unit; According to the feedback signal, set the reference voltage, forward integration period, reverse integration period, and integration time; According to the reference voltage, forward integration period, reverse integration period, and integration time, adopt the pixel-level bidirectional integration method to filter out the background information and output several pixel information of the moving target information within the window.

5. The intelligent air and space moving target perception system under a new type of complex background conditions according to claim 3, characterized in that: The pixel-level unidirectional integration method is specifically as follows: Step S31: Start the pixel-level unidirectional integration mode; Step S32: Set the reference voltage and integration time by the control signal generated by the information intelligent processing and feedback module; Step S33: Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 cumulative count; Step S34: Judge whether the integration time is over. If not, repeat Step S33. If so, output the total integration times.

6. The intelligent air and space moving target perception system under a new type of complex background conditions according to claim 4, characterized in that: The adoption of the pixel-level bidirectional integration method is specifically as follows: Step S41: Start the pixel-level bidirectional integration mode; Step S42: Set a reference voltage, a forward integration period, a reverse integration period, and an integration time; the integration time satisfies the condition that: where t int is the integration time, v is the target motion speed, and GSD is the pixel spatial resolution; Step S43: Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 forward count; Step S44: Judge whether the forward integration period is over. If not, repeat Step S43. If so, enter Step S45; Step S45: Perform reverse integration. Accumulate and integrate the charges generated by photoelectric conversion. When the integration voltage reaches the designed reference voltage, perform 1 reverse count; Step S46: Judge whether the reverse integration period is over. If not, repeat Step S45. If so, enter Step S47; Step S47: Judge whether the integration time is over. If not, repeat Steps S43 - S46. If so, record the difference between the forward count and the reverse count.

7. A space-air moving target intelligent perception system under a new type of complex background condition according to claim 1, characterized in that: The judgment and processing of several pixel information according to the working mode signal, and output the full-frame information or the moving target information within the window is specifically as follows: When the working mode signal is the target and background characteristic acquisition mode, judge whether several pixel information meet the output requirements. If so, perform data arrangement on several pixel information and output the full-frame information; if not, adjust the output feedback signal; When the working mode signal is the characteristic target detection mode, judge whether several pixel information meet the output requirements. If so, perform data arrangement on several pixel information and output the moving target information within the window; if not, adjust the output feedback signal.

8. A space-air moving target intelligent perception system under a new type of complex background conditions according to claim 7, characterized in that: The data arrangement of several pixel information is specifically as follows: The pixel information includes the spatial position of the information pixel generated by the pixel and the time characteristics of pixel sampling. According to the spatial position and sampling time characteristics of each pixel, arrange the pixel information in a predefined format and output the full-frame information or the moving target information within the window.

Images