Method and medium for removing motor interference in spaceborne limb scanning spectral imaging CCD

Abstract

A method and medium for removing motor interference in a spaceborne limb scanning spectral imaging CCD of the present invention processes the CCD digital image. Parameters such as the PWM driving frequency of the scanning motor, the pixel readout frequency of the CCD, and the pixel interference frequency are used in the processing. Starting from the CCD dark reference column, the interfered pixels are searched, and the purpose of eliminating all interference signals is achieved through the weighted linear interpolation and compensation method. The PWM wave driving frequency is cleverly utilized, combined with the CCD dark reference column and the pixel readout time, to locate the position of the interference pixel by pixel, eliminate the interference signal and then perform waveform fitting. This method is a preprocessing of digital images and is suitable for fast ground data processing; at the same time, it occupies less hardware logic resources and can also be used as a supplement to the on-board fast recognition algorithm, and can be applied to the fast recognition and decision-making of hot spots. Especially in the on-orbit flight stage of the payload, when the hardware shielding measures fail due to environmental problems, it can be used as an effective supplementary means.

Description

Technical Field

[0001] The present invention relates to the technical field of spaceborne limb scanning spectral imaging CCD, and particularly relates to a method and medium for removing motor interference in a spaceborne limb scanning spectral imaging CCD. Background Art

[0002] The limb payload carried on an atmospheric sounding satellite usually uses a scanning motor plus a CCD imaging detector to obtain the spatio-temporal distribution of trace gases in the direction of the stratospheric atmospheric cut height. The payload generally operates at an orbital altitude of 705 km. Using the scanning motor, the atmospheric measurement within the field of view is completed from bottom to top by scanning along the backward tangent direction. The normal working scanning range of the motor is 10 - 50 km, and it walks 60” (unit: second, 0.0165 degrees) per step, corresponding to a measurement height of 1.8 km. Then, within a range of 40 km, it walks about 22 steps per step. After the measurement is completed, it rotates back to zero and waits for the next measurement to start. The time for the motor to walk 22 steps plus the time to return to zero generally takes within 1 minute. To ensure the measurement efficiency, the single-step arrival time of the motor is usually controlled within 50 ms (millisecond), and the final return-to-zero time is within 3 s (second); to ensure the accuracy of the scanning area coverage, the pointing stability deviation of the limb scanning motor does not exceed 10” after each 60” stroke of a single step. To meet such high index requirements, a brushless DC motor is usually selected in the design. This type of motor has three-phase coils A / B / C and is star-connected. As Figure 1 shown in the equivalent circuit diagram of the scanning motor, the coils are divided into three parts, which are respectively connected to a drive bridge composed of MOS transistors (Metal Oxide Semiconductor field effect transistors). The parameters of the scanning motor in the figure are selected according to the requirements of the limb payload for the load moment of inertia, power consumption, and volume. In the control of the scanning motor, a three-closed-loop PID (Proportional Integral Derivative) algorithm of speed, position, and current is used to generate a PWM (Pulse Width Modulation) wave, which enters the three-phase MOS drive bridge for driving to complete the high-precision rotation and pointing stability of the motor.

[0003] However, this design brings inevitable interference problems at the same time. Figure 1In the [device], it is connected by a cable between the bridge drive circuit and the scanning motor. Cables with different wire diameters and lengths will generate parasitic inductances ranging from several μH (microhenries). The motor coil is wound with enameled wire and will generate parasitic capacitance. The combination of this parasitic capacitance and inductance will generate resonance at the moment of the PWM wave switching in the bridge circuit, as shown in Channel 2 of the oscilloscope screenshot in Figure 2. Channel 2 is the current waveform on the cable connecting the scanning motor, and the other channels are the three-phase drive waveforms of the scanning motor. The resonance signal will interfere with the CCD imaging signal in the ways of conduction and radiation.

[0004] Figure 2b It can be seen that after the rising edge of the PWM wave, the H-bridge MOSFET is turned on, and part of the drive current generates a damped attenuation waveform with a period of about 150 nS (nanoseconds). The attenuation period and amplitude are related to the parasitic capacitance of the motor coil, the parasitic inductance of the connecting cable, and the cable impedance. And the current waveform will interfere with the peripheral devices through channels such as the cable and the ground wire, especially devices like CCDs with deep potential wells, high dynamic ranges, and sensitivity to noise. As Figure 3 shown is the graph of the motor interference on the output of the edge-scanning spectroscopic imaging CCD displayed on the ground inspection platform. Figure 3 In [figure], the black area is the CCD imaging photo with a size of 1072 pixels X 258 pixels. The DN value curve part in the figure shows the DN values (Digital Number) of a selected row (1072 pixels) of pixels. It can be seen that in addition to the background noise, spike pulses appear regularly after every few pixels. The amplitudes of these spike pulses are more than 3 times the DN value of the background noise. These noises are the interference noises generated by the scanning motor.

[0005] Electromagnetic interference and related noise sources are the most difficult noise problems to solve in a CCD camera system. In engineering, means such as decoupling, conductor shielding, structural shielding, and grounding are usually used for the noise generated by electromagnetic interference. The decoupling network is used to remove the interference frequencies within the passband of CDS (Correlated Double Sampling). On the circuit board, the RC network is placed as close as possible to the power supply lines of the CCD and sensitive analog circuits. Many CCDs are installed on structural components due to heat dissipation reasons. Then, the drive lines on the circuit board are connected to the CCD pins through jump wires. The wires entering and leaving the CCD pins are also vulnerable to electromagnetic field coupling, introducing external noise. To solve this type of interference, the wires should be made as short as possible, and a grounded loop shield or twisted signal lines should be used. The shielding structure for the CCD and sensitive electronic devices is an effective means to reduce EMI (Electromagnetic Interference) noise problems. The shielding layer will reflect the incident EMI waves and absorb the unreflected part. The absorption loss acts on the electric field and magnetic field in a similar way. The reflection loss is larger in the electric field but relatively smaller in the magnetic field. Therefore, the absorption loss is mainly relied on to shield the CCD and surrounding electronic devices. In the CCD circuit design, the design of a multi-point grounding system can also provide better circuit isolation and shielding effects, and distinguish between the power ground, signal ground, digital ground, and analog ground. Using a large plane copper pour for the ground can effectively reduce the ground wire return path.

[0006] Regarding the above-mentioned interference phenomena of the scanning imaging CCD, the means of software denoising are limited and there is no relevant report. But the usual method is that because the interference frequency points of the scanning motor can be predicted, the Fourier transform is used for the digital image to find the interference frequency, and then the relevant noise can be eliminated through digital band-pass filtering. However, in the spectral imaging CCD system, the fast-changing absorption peaks of the spectral signals are continuously Gaussian distributed, and the distribution frequency is close to the motor drive interference frequency. The filter will reduce the amplitude of the useful signal. Summary of the Invention

[0007] A method for removing motor interference in a spaceborne limb scanning spectral imaging CCD proposed by the present invention obtains the interference signal frequency and position through the CCD dark pixel and row signal readout time positioning, so as to achieve the purpose of removing the CCD image interference signal generated by the motor.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] A method for removing motor interference in a spaceborne limb scanning spectral imaging CCD includes:

[0010] It includes the following steps:

[0011] S1. Confirm the scanning motor drive frequency F MOTOR, the CCD pixel readout frequency F CCD and the interference waveform frequency F INTER , the motor drive frequency F MOTOR , the CCD pixel readout frequency F CCD is obtained from the design value, and the interference waveform frequency F INTER can be obtained by collecting with an oscilloscope;

[0012] S2. According to the motor drive frequency F MOTOR , the CCD pixel readout frequency F CCD , obtain the interference repetition period Collect the DN values of the interference pixels in a single row of the CCD image through ground inspection to confirm the number of pixels m occupied by one interference; m is a natural number and m ≤ 2;

[0013] S3. Starting from the empty pixels in each row of the CCD, find the serial number of the first interference characteristic peak pixel marked as X1 and the serial number of the second interference characteristic peak pixel marked as Y1;

[0014] S4. Mark all the pixels with serial numbers: X1 + F REP ×n - 1, X1 + F REP ×n, X1 + F REP ×n + 1, Y1 + F REP ×n - 1, Y1 + F REP ×n, Y1 + F REP ×n + 1 in this row of pixels, and set their DN values to zero; n = 0, 1, 2... is a natural number;

[0015] Let the DN values of the above-mentioned serial number pixels be expressed as:

[0016]

[0017] S5. Let the serial numbers of two adjacent pixels of the marked interference pixels be (X1 + F REP ×n - 2, X1 + F REP ×n + 2) and (Y1 + F REP ×n - 2, Y1 + F REP ×n + 2), and the DN values are expressed as:

[0018]

[0019] Then the DN value of the interference pixel is backfilled by weighted linear interpolation to obtain:

[0020]

[0021] S6. Repeat steps S3 - S5 to obtain the entire interference-removed CCD image.

[0022] On the other hand, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to execute the steps of the above method.

[0023] As can be seen from the above technical solutions, the traditional measures for reducing EMI electromagnetic interference mainly adopt relevant means of software and hardware design. The method for removing motor interference in the on-board limb scanning spectral imaging CCD of the present invention intends to process the CCD digital image in a software manner, using parameters such as the PWM driving frequency of the scanning motor, the pixel readout frequency of the CCD, and the pixel interference frequency. Starting from the CCD dark reference column, the interfered pixels are searched, and through the weighted linear interpolation backfilling method, the purpose of eliminating all interference signals is achieved.

[0024] The optimized design of the hardware can only reduce the amplitude of the interference signal and cannot completely eliminate it. The present invention cleverly utilizes the PWM wave driving frequency, combines the CCD dark reference column and the pixel readout time, locates the interference position pixel by pixel, eliminates the interference signal and then performs waveform fitting to achieve the purpose of eliminating interference. This method is a preprocessing of digital images, with simple program design, suitable for fast ground data processing; at the same time, it occupies less hardware logic resources and can also be used as a supplement to the on-board fast recognition algorithm, and can be applied to the fast recognition and decision-making of hot spots. Especially in the case where the hardware shielding measures fail due to environmental problems during the on-orbit flight stage of the payload, it can be used as an effective supplementary means. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is the equivalent circuit diagram of the scanning motor drive;

[0026] Figure 2 is the resonant waveform after PWM switching of this drive axle; Figure 2a is the multi-period PWM waveform and current waveform; Figure 2b The rising edge of the single-period PWM waveform and the current waveform;

[0027] Figure 3 is the graph of the scanning imaging CCD being interfered;

[0028] Figure 4 is the timing relationship diagram of the scanning motor drive interference and CCD readout;

[0029] Figure 5 is the DN value situation of the first 60 pixels when the CCD reads one row;

[0030] Figure 6 is the DN value situation of the first 120 - 340 pixels when the CCD reads one row. DETAILED DESCRIPTION OF THE INVENTION

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0032] In the current scanning imaging CCD system, the existence of motor interference is an inevitable problem. In engineering, through careful circuit design and using a tight structure to wrap the motor, cables, and connectors, the leakage path of EMI interference sources is reduced. However, in actual engineering, the hardware design will bring many other problems. Reducing the debugging of the interference circuit will lead to an extension of the design cycle, and confirming the motor interference path and designing a tightly wrapped structural component will lead to an increase in cost. Especially in the on-board system, the strict requirements for weight and the single-unit envelope size often conflict with the on-board structure requirements. Of course, the EMI circuit and structural design measures without detailed modeling design can control most of the interference, but due to the complex on-board environment and the harsh requirements of in-orbit flight, it cannot be guaranteed that the payload scanning CCD system will not generate serious interference due to environmental changes during its life cycle, resulting in mission failure.

[0033] Since the frequencies of the interference signal and the spectral signal are similar, it is not easy for the software filtering algorithm to distinguish them; the complex filtering and multi-scale analysis techniques based on image processing may be effective, but the processing of a large amount of image data will lead to a reduction in computing efficiency, and the processing results are often released only after several weeks, which is not applicable to the current rapid early warning and hotspot positioning of spatio-temporal pollution.

[0034] The optimized design of the hardware can only reduce the amplitude of the interference signal and cannot completely eliminate it. The embodiments of the present invention cleverly utilize the PWM wave driving frequency, combine the CCD dark reference column and the pixel readout time, locate the position of the interference pixel by pixel, eliminate the interference signal and then perform waveform fitting to achieve the purpose of eliminating interference. This method is a preprocessing of digital images, with simple program design, suitable for rapid ground data processing; at the same time, it occupies less hardware logic resources and can also be used as a supplement to the on-board rapid recognition algorithm, and can be applied to the rapid recognition and decision-making of hotspot areas. Especially in the stage of the payload in-orbit flight, when the hardware shielding measures fail due to environmental problems, it can be used as an effective supplementary means.

[0035] The following is a specific description:

[0036] Based on Figure 4 a detailed elaboration of the technical solution, Figure 4 Figure 1 is a diagram showing the relationship between the CCD readout timing and the scanning motor drive timing. This is the three-phase drive timing of the scanning motor. To ensure the drive efficiency, the PWM drive frequency is usually on the order of KHz. In the figure, the value is set to 30 KHz (period: 33.3 μS), and the duty cycle is 50%. After the CCD exposure is completed, the main control unit reads out the CCD imaging values row by row. A row of CCD pixels mainly consists of blank pixels, dummy pixels, and imaging pixels. Blank pixels and dummy pixels do not respond to light signals and are mainly used for image offset correction. Figure 4 Among them, the number of blank pixels is 8, and the number of dummy pixels is 16. The readout frequency of each pixel is 751 KHz (period: 1.33 μS. To reduce the readout noise of the CCD, the pixel readout frequency is usually between 100 KHz and 1 MHz, and the readout value is designed according to the application scenario). That is to say, for every 25 pixels read out, there must be two interference signals (the rising and falling edges of the PWM wave will both generate interference) superimposed on these 25 pixels. Since the readout frequency and the motor drive frequency are not synchronized, the pixel positions where the interference signals are located are not fixed, but the interval is periodic with 25 pixels because the pixel readout frequency and the PWM drive frequency are fixed and unchanged. Figure 5 and Figure 6 respectively show the readout situation of the first 60 pixels and the readout situation of the middle imaging pixels. The distinction between interference waveform points is marked with an underline. As shown in the figure, for every 25 pixels, there must be an interference waveform. The marked points with an underline and the marked points without an underline respectively correspond to the interference signals generated by the rising and falling edges of the PWM wave. The readout of other rows is the same as this.

[0037] According to Figure 2b and Figure 4 the interference waveforms shown, its period is about 150 nS, and the interference time width is much smaller than the CCD single-pixel readout time of 1.33 μS. Therefore, on the readout pixel DN values, the abnormal points of the interference waveform are usually 1 - 2, and will not exceed 2 consecutive points, as Figure 5 and Figure 6 shown.

[0038] Based on the above characteristics, after confirming the scanning motor drive frequency, pixel readout frequency, and interference frequency, starting from the blank pixels of each row of the CCD, find the serial number of the first interference characteristic peak pixel marked as X1, and the serial number of the second interference characteristic peak pixel marked as Y1. Then the serial numbers of all pixels in this row are:

[0039] X1 + 25n - 1, X1 + 25n, X1 + 25n + 1;

[0040] Y1 + 25n - 1, Y1 + 25n, Y1 + 25n + 1; The DN values of the pixels are all set to 0, where n = 0, 1, 2... are natural numbers. It should be noted that starting from the blank pixels to find the interference characteristic peak can avoid the problem of finding the wrong peak due to the change of the spectral signal.

[0041] Let the DN values of the pixels with serial numbers X1 + 25n - 1, X1 + 25n, X1 + 25n + 1, Y1 + 25n - 1, Y1 + 25n, Y1 + 25n + 1 be expressed as:

[0042] D X1+25n-1 、D X1+25n 、D X1+25n+1 、D Y1+25n-1 、D Y1+25n 、D X1+25n+1

[0043] Since the DN values of the spectral signals do not change drastically between adjacent pixels, let the serial numbers of two adjacent pixels of the interfering pixel be (X1 + 25n - 2, X1 + 25n + 2) and (Y1 + 25n - 2, Y1 + 25n + 2). Then, the DN value of the interfering pixel is interpolated and compensated back through weighted linear interpolation to obtain:

[0044]

[0045] After that, each row of the CCD is processed in the above manner to obtain a complete CCD image free of interference.

[0046] The above values based on Figure 4 the described method for processing the interference of the scanning imaging CCD motor are not limited to the above specific values. The method is summarized as follows:

[0047] 1. Confirm the driving frequency F of the scanning motor MOTOR 、the readout frequency F of the CCD pixels CCD and the interference waveform frequency F INTER . The driving frequency F of the motor MOTOR 、the readout frequency F of the CCD pixels CCD are obtained from the designed values, and the interference waveform frequency F INTER can be acquired by collecting with an oscilloscope;

[0048] 2. According to the driving frequency F of the motor MOTOR 、the readout frequency F of the CCD pixels CCD , obtain the interference repetition period Collect the DN values of the interfering pixels in a single row of the CCD through ground inspection to confirm the number m of pixels occupied by one interference; m is a natural number, and usually m ≤ 2;

[0049] 3. Starting from the empty pixels in each row of the CCD, find the serial number of the first interference characteristic peak pixel marked as X1, and the serial number of the second interference characteristic peak pixel marked as Y1;

[0050] 4. For all pixels in this row with serial numbers: X1 + F REP ×n - 1, X1 + F REP ×n, X1 + F REP×n + 1, Y1 + F REP ×n - 1, Y1 + F REP ×n, Y1 + F REP Mark the pixel of ×n + 1 and set its DN value to zero; n = 0, 1, 2… are natural numbers; Let the DN values of the above serial-numbered pixels be expressed as:

[0051]

[0052] 5. Let the serial numbers of two adjacent pixels of the marked interfering pixel be (X1 + F REP ×n - 2, X1 + F REP ×n + 2) and (Y1 + F REP ×n - 2, Y1 + F REP ×n + 2), and the DN values are expressed as:

[0053]

[0054] Then the DN value of the interfering pixel is backfilled through weighted linear interpolation to obtain:

[0055]

[0056] 6. Repeat steps 3 - 5 to obtain the entire CCD image with interference removed.

[0057] In summary, the embodiments of the present invention are intended to process CCD digital images in software. During the processing, parameters such as the PWM driving frequency of the scanning motor, the CCD pixel reading frequency, and the pixel interference frequency are used. Starting from the CCD dark reference column, the interfering pixels are found, and through the weighted linear interpolation backfilling method, the purpose of eliminating all interference signals is achieved.

[0058] The optimized design of the hardware can only reduce the amplitude of the interference signal and cannot completely eliminate it. The present invention cleverly utilizes the PWM wave driving frequency, combines the CCD dark reference column and the pixel reading time, locates the interference position pixel by pixel, eliminates the interference signal and then performs waveform fitting to achieve the purpose of eliminating interference. This method is a preprocessing of digital images, with simple program design, suitable for fast ground data processing; at the same time, it occupies less hardware logic resources and can also be used as a supplement to the on - satellite fast recognition algorithm, and can be applied to the fast recognition and decision - making of hot spots. Especially during the in - orbit flight stage of the payload, when the hardware shielding measures fail due to environmental problems, it can be used as an effective supplementary means.

[0059] On the other hand, the present invention also discloses a computer - readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is caused to execute the steps of any of the above - mentioned methods.

[0060] In another aspect, the present invention also discloses a computer device, including a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor is caused to execute the steps of any of the above methods.

[0061] In another embodiment provided by the present application, there is also provided a computer program product containing instructions. When it runs on a computer, the computer is caused to execute the steps of any of the above methods.

[0062] It can be understood that the system provided by the embodiments of the present invention corresponds to the method provided by the embodiments of the present invention. For the explanations, examples, and beneficial effects of related content, reference can be made to the corresponding parts in the above methods.

[0063] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0064] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0065] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for removing motor interference in a spaceborne limb-scanning spectral imaging CCD, characterized in that: including the following steps, S1. Confirm the scanning motor drive frequency F MOTOR , the CCD pixel readout frequency F CCD and the interference waveform frequency F INTER . The motor drive frequency F MOTOR , the CCD pixel readout frequency F CCD is obtained from the design values, and the interference waveform frequency F INTER can be acquired by oscilloscope sampling; S2. According to the motor drive frequency F MOTOR , and the CCD pixel readout frequency F CCD , obtain the interference repetition period Collect the DN values of the interference pixels in a single row of the CCD image through ground inspection to confirm the number of pixels m occupied by one interference; m is a natural number and m ≤ 2; S3. Starting from the empty pixels in each row of the CCD, find the serial number of the first interference feature peak pixel marked as X1 and the serial number of the second interference feature peak pixel marked as Y1; S4. Mark the pixels with all serial numbers of: X1 + F in this row REP ×n - 1, X1 + F REP ×n, X1 + F REP ×n + 1, Y1 + F REP ×n - 1, Y1 + F REP ×n, Y1 + F REP ×n + 1, and set their DN values to zero; n = 0, 1, 2... is a natural number; Let the DN values of the above serial number pixels be expressed as: S5. Let the serial numbers of two adjacent pixels of the marked interfering pixel be (X1 + F REP ×n - 2, X1 + F REP ×n + 2) and (Y1 + F REP ×n - 2, Y1 + F REP ×n + 2), and the DN value is expressed as: Then the DN value of the interference pixel is interpolated and filled back through weighted linear interpolation to obtain: S6. Repeat steps S3 - S5 to obtain the entire CCD image with interference removed.

2. A computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to perform the steps of the method according to claim 1.

Images