Quantitative evaluation method for image quality of space-based infrared imaging system

Abstract

The invention belongs to the technical field of infrared remote sensing and infrared space, and discloses a space-based infrared imaging system image quality quantitative evaluation method, which comprises the following steps: initializing imaging system optical and detector parameters; selecting an image quality quantitative evaluation index; performing forward processing on the evaluation indexes; constructing an infrared system image quality quantitative evaluation model based on a TOPSIS entropy weight method; selecting an optimal system parameter according to a model evaluation result; the method comprises the following steps: firstly setting detection parameters of an optical system and a detector system of an infrared detection system, then selecting an image quality quantification parameter of the infrared system, quantitatively evaluating the image quality of the infrared system through a TOPSIS entropy weight method, and selecting an optimal detection parameter according to an evaluation result; the invention provides an infrared imaging system image quality quantitative evaluation method, the detection parameters selected by the method can effectively improve the infrared image quality, and the method is suitable for the infrared technical field.

Description

Technical Field

[0001] The invention relates to the field of infrared remote sensing and infrared space technology, and in particular to a method for quantitatively evaluating image quality of a space-based infrared imaging system. Background Art

[0002] In the field of infrared remote sensing, traditional infrared system imaging evaluation algorithms usually only consider a single characteristic, using a single indicator such as modulation transfer function (MTF), noise equivalent temperature difference (NETD), and minimum resolvable temperature difference (MRTD) to evaluate the performance of simulation systems and real systems, which has become the standard for traditional photoelectric system performance evaluation. Traditional evaluation models are somewhat subjective, and the test results under different experimental conditions vary greatly. In addition, they do not take into account nonlinear response factors, making it difficult to support the evaluation of the overall system performance.

[0003] In modern payload design, many parameters are mutually constrained: pursuing weak target detection performance requires improving the sensitivity of the payload, which will lead to an increase in suspected targets and an increase in false alarm rate. Therefore, it is crucial to select appropriate quantitative evaluation indicators for infrared image quality.

[0004] The image evaluation system of the infrared imaging system starts with the specific detection task requirements. The diffraction limit GSD or angular resolution is a strong constraint indicator for the optical system design. Excessive high resolution requirements will lead to excessive optical aperture requirements, which will be difficult to achieve and further limit the optical field of view; optical quality factor Q, detection distance R, pixel size dx and cutoff wavelength λ max Together they determine the optical focal length f and the minimum aperture requirement D min and F # In terms of detector performance, the required sensitivity begins to shift from the background limit to a lower noise limit. Parameters such as the noise equivalent sensitivity coefficient, signal-to-noise ratio, detection spectrum, and minimum and maximum temperature detection range comprehensively determine the image quality of the imaging system.

[0005] Therefore, an intelligent evaluation algorithm with downward analysis capability is needed to effectively accelerate the iteration speed of the imaging system quality evaluation-system design loop, provide technical support for the research and development of on-orbit payloads, and is also of great significance in target detection, identification and tracking. Summary of the invention

[0006] The present invention aims to provide a method for quantitatively evaluating image quality of a space-based infrared imaging system to solve the problems described in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A method for quantitatively evaluating image quality of a space-based infrared imaging system comprises the following steps:

[0009] S101, initializing optical system parameters and detector parameters of the infrared imaging system;

[0010] Select the target and background, set the lighting conditions, spectrum and seasonal conditions, and load the scene radiance image. Set the infrared system parameters, where the optical system parameters include camera aperture D, focal length, detection distance, optical efficiency, central wavelength, energy concentration and instantaneous field of view. Then set the detector parameters, which are: pixel size, dark current, integration time, instrument background noise, readout noise, full well number of electrons, non-uniformity, number of quantization bits and quantum efficiency.

[0011] S102, selecting a detection system image quality evaluation index according to infrared imaging system parameters;

[0012] This invention mainly studies the method of establishing the image quality evaluation system of the infrared imaging system of aerial targets. First, the modules in the imaging link of the infrared system are analyzed. Starting from the overall indicators that need to be achieved for aerial target imaging, the indicators of the infrared detector level of the core component of the payload are decomposed level by level. The resolution, band, integration time and other indicators required for aerial target imaging are demonstrated, and the imaging quality analysis of aerial targets by typical infrared systems is further carried out. Therefore, the evaluation indicators selected by this method are: system F number, ground resolution, optical diffraction limit resolution, IFOV, minimum resolution angle, optical efficiency, energy concentration, quantum efficiency, signal electron number, noise electron number, and electron number signal-to-noise ratio.

[0013] S102.1. The core parameters of the optical system are determined based on the angular resolution and the optical diffraction limit.

[0014] First is the relationship between the camera angular resolution IFOV and the ground detection resolution GSD:

[0015]

[0016] GSD=IFOV*R

[0017] Where IFOV is the angular resolution, f is the focal length of the optical system, dx is the pixel size, GSD is the ground resolution, and R is the system detection distance;

[0018] The optical diffraction-limited resolution σ is:

[0019]

[0020] Where D is the aperture of the optical system, and λ is the detection wavelength;

[0021] Minimum resolution angle θ min for:

[0022]

[0023] System limit caliber D min for:

[0024]

[0025] Among them, F # is the system F number;

[0026] S102.2. The infrared system noise is evaluated through the electron number signal-to-noise ratio indicator. Starting from the internal detector level indicators, the impact of each noise sub-item such as photon noise, instrument noise, dark current noise and electronic noise on sensitivity and dynamic range is analyzed. Starting from the top-level signal-to-noise ratio indicator, considering the maximum and minimum levels of the observed signal, indicators are allocated to the lower-level noise items to reasonably evaluate the system imaging indicator requirements.

[0027] Signal electron number calculation. According to the signal electron number N s Definition: The number of charges generated by the target and injected into the integral capacitor during an integral time period. The number of electrons received by the target pixel is expressed as:

[0028]

[0029] Where QE is the detector quantum efficiency, λ is the central wavelength, and h is the Planck constant (h = 6.6 × 10 -34 J.s -1 ), c is the speed of light (c = 3 × 10 8 m / s), J is the target radiation intensity, τ a is the atmospheric transmittance at the target height, τ0 is the optical system transmittance (including optical shielding factor), EE is the energy concentration, l is the detection distance, T int is the system integration time.

[0030] The number of electrons N generated by observing the Earth background and its atmospheric path in addition to the target bg It is expressed as:

[0031]

[0032] Where L bg is the background radiance.

[0033] S102.3. In space-based long-distance imaging, the images of aerial targets in infrared systems are mostly point targets, which are less affected by spatial noise. The system noise can be simplified to temporal noise:

[0034]

[0035] n read is the number of read noise electrons in the system, n phThe RMS value produced by the sum of all photon noise correlations is:

[0036]

[0037] N instr It is the sum of the radiation of all components of the optical instrument and the Dewar window, N dark is the number of dark current noise electrons in the infrared detector.

[0038] In engineering, the noise performance of infrared focal plane is generally described by the electron number signal-to-noise ratio model, where the electron number signal-to-noise ratio is the ratio of the number of signal electrons to the total number of noise electrons.

[0039]

[0040] The upper limit of the system detection performance dominated by photon noise can be easily determined, but because the system also includes many factors such as detector noise, electronic readout noise, instrument background, etc., the above process is only used for the initial simplified model to evaluate the system capability.

[0041] S103, performing dimensionless normalization processing on the evaluation indicators respectively;

[0042] The image quality evaluation indexes of the selected infrared imaging system are positively processed.

[0043] Let the index be x, and the positive method of the extremely small index is:

[0044] posit xm =max(x)-x m ,m=1,2,3…

[0045] The positive method of very large indicators is:

[0046] posit xm =x m -min(x),m=1,2,3…

[0047] S104. Construct a quantitative evaluation model of infrared system image quality based on TOPSIS entropy weight method;

[0048] S104.1. Construct an evaluation matrix based on the quantitative evaluation index of image quality:

[0049]

[0050] Where n is the number of evaluation objects and m is the number of evaluation indicators. In order to remove the influence of dimension, X is normalized and recorded as Z:

[0051]

[0052] S104.2. Calculate the weight of each evaluation index using the entropy weight method and calculate the entropy value of each index:

[0053]

[0054] Where k is related to the number of samples, and is generally taken as k = 1 / ln(n). Then the weights of each indicator are calculated:

[0055]

[0056] S104.3. Determine the optimal and the worst distances. First, define the maximum and minimum values:

[0057] Z + =(max(z 11 ,z 21 ,...,z m1 ),max(z 12 ,z 22 ,...,z m2 ),...,max(z 1n ,z 2n ,...,z mn ))

[0058] Z - =(min(z 11 ,z 21 ,...,z m1 ),min(z 12 ,z 22 ,...,z m2 ),...,min(z 1n ,z 2n ,...,z mn ))

[0059] Calculate the optimal distance and the worst distance of the i-th evaluation object:

[0060]

[0061] S104.4. Calculate the scores of each group of parameters based on the obtained optimal and worst distances:

[0062]

[0063] S105. Implement quantitative evaluation of the image quality of the imaging system according to the model evaluation result.

[0064] The image quality of the imaging system is evaluated based on the normalized results of the model evaluation:

[0065]

[0066] The comprehensive ranking of imaging quality can be obtained based on the scores.

[0067] Principle and beneficial effects of the technical solution: The present invention comprehensively evaluates the performance indicators of each subsystem of the infrared detection system, and performs a performance analysis on the imaging system determined by the indicators. According to the task objectives, three levels of initialization and design are adopted: the first level initializes the infrared imaging system scene, the second level selects imaging quality evaluation indicators based on detector selection, scale and pixel size, and the third level calculates performance through comprehensive indicators. This method is objective and universal, and is of great significance for performance evaluation, system improvement, cost savings and decision support. Compared with the prior art, the present invention constructs an infrared system imaging quality evaluation indicator, and combines the entropy weight TOPSIS calculation to obtain the weight value of each indicator in the imaging quality evaluation. Finally, based on the weight index of each indicator, the evaluation of multiple groups of detection system parameters is realized, which has a certain effectiveness and provides reliable tools and methods for the application and development of infrared technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 A flowchart of a method for quantitatively evaluating image quality of a space-based infrared imaging system according to the present invention;

[0069] Figure 2 The relationship between the core parameters of the infrared imaging system and the system performance;

[0070] Figure 3 A flow chart of a method for quantitatively evaluating image quality of an infrared imaging system provided by an embodiment of the present invention;

[0071] Figure 4 Initialize an infrared radiance image of a scene for an embodiment of the present invention;

[0072] Figure 5 The imaging results of multiple sets of detection system parameters according to the embodiment of the present invention are shown; DETAILED DESCRIPTION

[0073] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:

[0074] S101, initializing optical system parameters and detector parameters of the infrared imaging system;

[0075] In this embodiment, a space-based medium-wave infrared detection scene is selected, the target is an aircraft target, the background is a sea of ​​clouds, the illumination condition is set to a zenith angle of 20 degrees, the season condition is summer, and the scene radiance image is loaded, such as Figure 4As shown. Set the infrared system parameters, where the optical system parameters include camera aperture D, focal length, track height, optical efficiency, central wavelength, energy concentration, instrument background noise and instantaneous field of view. Then set the detector parameters, which are: pixel size, dark current, integration time, readout noise, full well electron number, non-uniformity, quantization bit number and quantum efficiency.

[0076] S102. Select imaging quality evaluation indicators based on infrared imaging system parameters. The infrared detection imaging process is affected by the optical system, detector, and other external stray light. The evaluation indicators selected by this method are mainly the core parameters of the optical system and the core parameters of the detector system, including: system F number, optical diffraction limit, angular resolution, optical efficiency, energy concentration, integration time, quantum efficiency, signal electron number, noise electron number, and electron number signal-to-noise ratio. There are a total of 10 indicator parameters, see Table 1 below for details:

[0077] Table 1. Quantitative evaluation system for image quality of infrared imaging system

[0078]

[0079]

[0080] S103, performing dimensionless normalization processing on the image quality evaluation indexes respectively;

[0081] Let the index be x, and the positive method of the extremely small index is:

[0082] posit xm =max(x)-x m ,m=1,2,3…

[0083] The positive method of very large indicators is:

[0084] posit xm =x m -min(x),m=1,2,3…

[0085] S104. Construct a quantitative evaluation model of infrared system image quality based on TOPSIS entropy weight method;

[0086] S104.1. Construct an evaluation matrix based on the quantitative evaluation index of image quality:

[0087]

[0088] Where n is the number of evaluation objects and m is the number of evaluation indicators. In order to remove the influence of dimension, X is normalized and recorded as Z:

[0089]

[0090] S104.2. Calculate the weight of each evaluation index using the entropy weight method and calculate the entropy value of each index:

[0091]

[0092] Where k is related to the number of samples, and is generally taken as k = 1 / ln(n). Then the weights of each indicator are calculated:

[0093]

[0094] S104.3. Determine the optimal and the worst distances. First, define the maximum and minimum values:

[0095] Z + =(max(z 11 ,z 21 ,...,z m1 ),max(z 12 ,z 22 ,...,z m2 ),...,max(z 1n ,z 2n ,...,z mn ))

[0096] Z - =(min(z 11 ,z 21 ,…,z m1 ),min(z 12 ,z 22 ,...,z m2 ),...,min(z 1n ,z 2n ,...,z mn ))

[0097] Then calculate the optimal distance and the worst distance of the i-th evaluation object:

[0098]

[0099] S104.4. Calculate the scores of each group of parameters based on the obtained optimal and worst distances:

[0100]

[0101] S105, quantitatively evaluate the image quality of the imaging system according to the model evaluation results. Perform imaging evaluation of the imaging system according to the normalized results of the model evaluation:

[0102]

[0103] The comprehensive ranking of detection performance can be obtained based on the scores.

[0104] The following are multiple groups of infrared imaging detection system parameters to be evaluated in this embodiment.

[0105] Table 2. Initial parameters of multiple groups of infrared imaging detection systems to be evaluated

[0106] parameter System 1 System 2 System 3 System 4 System 5 System 6 Diameter(m) 0.5 0.55 0.4 0.5 0.4 0.6 Focal length(m) 1.2 1.12 1.2 1.16 1.18 1.3 Detection distance(Km) 6000 6000 6000 6000 6000 6000 Central wavelength (μm) 4 4 4 4 4 4 Integration time (ms) 1 1 1 1 1 1 Pixel size (μm) 15 16 18 16 17 16 Readout noise (e-) 600 600 600 600 600 600 Full well number of electrons (Me-) 6 6 6 6 6 6 Quantization bit number (bit) 12 12 12 12 12 12

[0107] Table 3. Evaluation indicators of multiple groups of infrared imaging detection systems to be evaluated

[0108]

[0109]

[0110] Table 4. Weight values ​​of each indicator

[0111]

[0112] Table 5. System evaluation score results and ranking

[0113]

[0114]

[0115] We call images and videos of visual environments acquired using high-quality capture devices (infrared cameras, cameras, etc.) that can operate within the spectral range a natural scene. To verify the effectiveness of the present invention, its evaluation results are verified with the widely recognized natural image evaluation model No Reference Image Quality Evaluation (Natural Image Quality Evaluator, NIQE) and the improved "Completely Blind" Image Quality Evaluation Model (Integrated Local NIQE, IL-NIQE). (The lower the evaluation scores of NIQE and IL-NIQE, the better the image quality).

[0116] Table 6. System evaluation scores and rankings

[0117]

[0118] This result can be compared with Figure 5 The simulation imaging results of multiple sets of detection system parameters are compared and verified. Figure 5 It can be seen that:

[0119] (1) System 6 and System 2 have relatively good subjective effects, while System 4, System 5, System 1 and System 3 have relatively large noise fluctuations and poor image quality. This evaluation method is consistent with subjective judgment.

[0120] (2) Further analysis of system 6 and system 2 shows that in scenes containing targets, the signal-to-noise ratio of system 6 is greater than that of system 2, indicating that the imaging quality of system 6 is greater than that of system 2. The F number of system 6 is smaller than that of system 2, and the light-gathering ability of system 6 is stronger. In addition, the optical diffraction resolution and minimum resolution angle of system 6 are better than those of system 2, and the system has a stronger detection spatial range. Therefore, the overall imaging quality of system 6 is better than that of system 2.

[0121] (3) From the clutter fluctuations, it can be seen that the clutter noise of system 4 is smaller than that of system 5. Combining parameters such as SNR, GSD, and IFOV, it can be concluded that the parameter configuration of imaging system 4 is better than that of system 5.

[0122] (4) Subjectively, both system 5 and system 1 have stronger noise. System 5 has higher target energy concentration and smaller energy divergence, which is more suitable for practical target detection applications. Combining evaluation indicators such as EE, OE, and QE, it can be concluded that the imaging quality of system 5 is better than that of system 1.

[0123] (5) Based on human vision, the imaging noise fluctuation of system 3 is the highest and the quality is the worst.

[0124] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the technical solution of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for quantitatively evaluating image quality of a space-based infrared imaging system, characterized in that: The following steps are involved: S101, initializing optical system parameters and detector parameters of the infrared imaging system; S102, selecting a detection system image quality evaluation index according to infrared imaging system parameters; S103, performing dimensionless normalization processing on the evaluation indicators respectively; S104. Construct a quantitative evaluation model of infrared system image quality based on TOPSIS entropy weight method; S105. Evaluate the image quality of the detection system according to the model evaluation result.

2. The method for quantitatively evaluating image quality of a space-based infrared imaging system according to claim 1, characterized in that: In S101, the initialized optical system parameters include: camera aperture D, focal length, detection distance, optical efficiency, central wavelength, energy concentration and instantaneous field of view; the initialized detector parameters include: pixel size, dark current, integration time, instrument background noise, readout noise, full well electron number, non-uniformity, quantization bit number and quantum efficiency.

3. The method for quantitatively evaluating image quality of a space-based infrared imaging system according to claim 2, characterized in that: In S102, the image quality evaluation index of the infrared imaging system is selected. The selected evaluation indexes are the core parameters of the optical system and the core parameters of the detector system, including: system F number, ground resolution, optical diffraction limit resolution, instantaneous field of view (IFOV), minimum resolution, optical efficiency, energy concentration, quantum efficiency, signal electron number, noise electron number, and electron number signal-to-noise ratio. The calculation of some evaluation indexes is as follows: S102.

1. The core parameters of the optical system are determined by the angular resolution and the optical diffraction limit. First, the relationship between the camera's instantaneous field of view IFOV and the ground detection resolution GSD is: GSD=IFOV*R Where f is the focal length of the optical system, dx is the pixel size, GSD is the ground resolution, and R is the system detection distance; The optical diffraction-limited resolution σ is: Where D is the aperture of the optical system, and λ is the detection wavelength; Minimum resolution angle θ min for: System limit caliber D min for: Among them, F # is the system F number; S102.

2. Calculation of signal electron number, signal electron number N s The definition of is: the number of charges generated by the target and injected into the integral capacitor within an integral time period, and the number of electrons received by the target pixel, which can be expressed as: Where QE is the detector quantum efficiency, λ is the central wavelength, and h is the Planck constant (h = 6.6 × 10 -34 J.s -1 ), c is the speed of light (c = 3 × 10 8 m / s), J is the target radiation intensity, τ0 is the atmospheric transmittance at the target height, τ0 is the optical system transmittance (including optical shielding factor), EE is the energy concentration, R is the detection distance, T int is the system integration time; The number of electrons N generated by observing the Earth background and its atmospheric path in addition to the target bg It is expressed as: Where L bg is the background radiance. S102.

3. Calculation of noise electron count: In space-based long-distance imaging, the image of an aerial target in the infrared system is mostly a point target, which is less affected by spatial noise. The system noise can be simplified to temporal noise: n read is the number of read noise electrons in the system, n ph The RMS value produced by the sum of all photon noise correlations is: N instr It is the sum of the radiation of all components of the optical instrument and the Dewar window, N dark is the number of dark current noise electrons in the infrared detector.

4. The method for quantitatively evaluating image quality of a space-based infrared imaging system according to claim 3, characterized in that: In S103, the selected evaluation index of the infrared imaging system is subjected to index attribute positive processing: Let the index be x, and the positive method of the extremely small index is: posit xm =max(x)-x m ,m=1,2,3… The positive method of very large indicators is: posit xm =x m -min(x),m=1,2,3… 5. The method for quantitatively evaluating image quality of a space-based infrared imaging system according to claim 4, characterized in that: In S104, a quantitative evaluation model of infrared system image quality based on TOPSIS entropy weight method is constructed; S104.

1. Construct an evaluation matrix based on the quantitative evaluation index of image quality: Where n is the number of evaluation objects, m is the number of evaluation indicators, and in order to remove the influence of dimension, X is normalized and recorded as Z: S104.

2. Calculate the weight of each evaluation index using the entropy weight method and calculate the entropy value of each index: Among them, k is related to the number of samples, generally k = 1 / ln(n), and then the weights of each indicator are calculated: S104.

3. To determine the optimal and the worst distances, first define the maximum and minimum values: WITH + =(max(of 11 ,With 21 ,…,With m1 ),max(of 12 ,With 22 ,...,With m2 ),...,max(of 1n ,With 2n ,...,With mn )) WITH - =(min(of 11 ,With 21 ,...,With m1 ),min(of 12 ,With 22 ,...,With m2 ),...,min(z 1n ,With 2n ,...,With mn )) Then calculate the optimal distance and the worst distance of the i-th evaluation object: S104.

4. Calculate the scores of each group of parameters using the obtained optimal and worst distances:

6. The method for quantitatively evaluating image quality of a space-based infrared imaging system according to claim 5, characterized in that: In S105, the image quality of the infrared imaging system is evaluated according to the normalized result of the model evaluation: Based on the scores, we can get a comprehensive ranking of the infrared imaging system’s image performance.

Images