Multi-band fusion image quality evaluation method and device

Abstract

The invention provides a multi-band fusion image quality evaluation method and device, relates to the technical field of aviation systems, and is used for improving the accuracy of multi-band fusion image quality evaluation so as to reflect the visual perception effect of a pilot in actual flight. The method mainly comprises the following steps: obtaining subjective evaluation scores and objective evaluation scores of a multi-band fusion image obtained through a plurality of fusion algorithms; calculating a correlation coefficient according to the subjective evaluation score and the objective evaluation score; determining a forward objective evaluation index corresponding to each evaluation dimension from objective evaluation indexes for evaluating the EVS application scene under different evaluation dimensions through the correlation coefficient; and calculating an image quality score of the multiband fusion image according to the subjective evaluation score and the forward objective evaluation index corresponding to each evaluation dimension.

Description

Technical Field

[0001] The present application relates to the technical field of avionics systems, and particularly to a multi-band fusion image quality evaluation method and device. Background Art

[0002] The evaluation of image fusion quality is particularly important in a multi-band enhanced vision system (EVS). The quality of the image directly affects the pilot's decision-making and safe operation. The image fusion involved in EVS usually includes images from different bands such as visible light and infrared. Each band of images has different resolutions, information contents, and noise characteristics. For example, infrared images can provide visible information under low light conditions, but their resolutions are relatively low; visible light images provide high resolutions and clear details, but perform poorly at night and in bad weather. How to balance the advantages and disadvantages of these different band images and ensure that the fused image is both clear and contains rich information is a key issue in image fusion quality assessment.

[0003] Currently, the evaluation methods of image fusion quality mainly rely on objective evaluation indicators in various dimensions such as visual fidelity, information entropy, contrast and detail features, structural similarity, and correlation. However, they all have limitations and cannot prove that they truly reflect the fusion effect. Especially in the application scenario of EVS, relying solely on objective evaluation often cannot fully reflect the visual perception effect of the pilot during actual flight. Summary of the Invention

[0004] Embodiments of the present application provide a multi-band fusion image quality evaluation method and device, which are used to improve the accuracy of multi-band fusion image quality evaluation, so as to reflect the visual perception effect of the pilot during actual flight.

[0005] Embodiments of the present invention provide a multi-band fusion image quality evaluation method, and the method includes:

[0006] Obtain the subjective evaluation scores and objective evaluation scores of multi-band fusion images obtained by various fusion algorithms respectively;

[0007] Calculate the correlation coefficient according to the subjective evaluation score and the objective evaluation score;

[0008] Determine the positive objective evaluation indicators corresponding to each evaluation dimension from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient;

[0009] Calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension.

[0010] In an alternative embodiment, obtaining the subjective evaluation scores and objective evaluation scores of the multi-band fusion images obtained by a variety of fusion algorithms respectively includes:

[0011] Performing fusion on the sample images in the test dataset through the fusion algorithm to be evaluated and the benchmark fusion algorithm respectively to obtain the multi-band fusion image to be evaluated and the benchmark multi-band fusion image;

[0012] Determining the subjective evaluation score of each multi-band fusion image to be evaluated according to the evaluation score of the multi-band fusion image to be evaluated relative to the benchmark multi-band fusion image by the tester;

[0013] Evaluating the multi-band fusion image to be evaluated through a variety of objective evaluation metrics to obtain the objective evaluation scores corresponding to the variety of objective evaluation metrics respectively.

[0014] In an alternative embodiment, determining the positive objective evaluation metric corresponding to each evaluation dimension from the objective evaluation metrics for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient includes:

[0015] Among the objective evaluation metrics for evaluating the EVS application scenario under different evaluation dimensions, selecting the top N objective evaluation metrics with the correlation coefficient as the positive objective evaluation metric corresponding to the corresponding evaluation dimension.

[0016] In an alternative embodiment, calculating the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation metric corresponding to each evaluation dimension includes:

[0017] Obtaining the Spearman rank correlation coefficient of the positive objective evaluation metric corresponding to each evaluation dimension, and calculating the comprehensive objective evaluation score according to the Spearman rank correlation coefficient of the positive objective evaluation metric corresponding to each evaluation dimension;

[0018] Performing weighted calculation on the subjective evaluation score and the comprehensive objective evaluation score to obtain the image quality score of the multi-band fusion image.

[0019] In an alternative embodiment, calculating the comprehensive objective evaluation score according to the Spearman rank correlation coefficient of the positive objective evaluation metric corresponding to each evaluation dimension includes:

[0020] Obtaining the weight value of the positive objective evaluation metric of each evaluation dimension;

[0021] Performing weighted calculation on the weight values and positive objective evaluation metrics of the positive objective evaluation metrics of all evaluation dimensions to obtain the comprehensive objective evaluation score of each evaluation dimension.

[0022] In an alternative embodiment, the method further includes:

[0023] Determine the initial weight value of the positive objective evaluation index in the corresponding evaluation dimension according to the ratio of the correlation coefficient of the positive objective evaluation index to the total score of the correlation coefficients of all positive objective evaluation indexes in the evaluation dimension where it is located.

[0024] In an alternative embodiment, the method further includes:

[0025] Update the weight value of the positive objective evaluation index by using the particle swarm filtering algorithm.

[0026] An embodiment of the present invention provides a multi-band fusion image quality evaluation device, and the device includes:

[0027] An acquisition module, configured to acquire the subjective evaluation score and the objective evaluation score of the multi-band fusion image obtained by using multiple fusion algorithms respectively;

[0028] A calculation module, configured to calculate the correlation coefficient according to the subjective evaluation score and the objective evaluation score;

[0029] A determination module, configured to determine the positive objective evaluation index corresponding to each evaluation dimension from the objective evaluation indexes of the evaluation EVS application scenario in different evaluation dimensions through the correlation coefficient;

[0030] The calculation module is further configured to calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation index corresponding to each evaluation dimension.

[0031] A computer program product, the computer program product includes a computer program, and when the computer program is executed by a processor, it implements the above multi-band fusion image quality evaluation method.

[0032] The present invention provides a multi-band fusion image quality evaluation method and device. First, subjective evaluation scores and objective evaluation scores of multi-band fusion images obtained by various fusion algorithms are acquired. Then, a correlation coefficient is calculated based on the subjective evaluation scores and the objective evaluation scores. After that, positive objective evaluation indicators corresponding to each evaluation dimension are determined from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient. Finally, an image quality score of the multi-band fusion image is calculated based on the subjective evaluation scores and the positive objective evaluation indicators corresponding to each evaluation dimension. Since the image quality score of the multi-band fusion image in the present application comprehensively considers the subjective evaluation scores and the positive objective evaluation indicators, and the positive objective evaluation indicator can represent the consistency between the objective indicator and the subjective evaluation, the image quality score of the multi-band fusion image obtained through the present application can reflect the visual perception effect of the pilot during actual flight. Furthermore, the accuracy of multi-band fusion image quality evaluation can be improved through the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flowchart of a multi-band fusion image quality evaluation method provided by the present application;

[0034] Figure 2 It is a flowchart of a method for obtaining subjective evaluation scores provided by the present application;

[0035] Figure 3 It is a schematic diagram of a correlation coefficient calculation process provided by the present application;

[0036] Figure 4 It is a schematic diagram of objective evaluation indicators corresponding to an evaluation dimension provided by the present application;

[0037] Figure 5 It is a schematic diagram of the principle of multi-band fusion image quality evaluation provided by the present application;

[0038] Figure 6 It is a schematic structural diagram of a multi-band fusion image quality evaluation device provided by the present application;

[0039] Figure 7 It is a schematic diagram of a computer device provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to better understand the above technical solutions, the technical solutions of the embodiments of the present application will be described in detail below through the accompanying drawings and specific embodiments. It should be understood that the specific features in the embodiments of the present application and the embodiments are detailed descriptions of the technical solutions of the embodiments of the present application, rather than limitations on the technical solutions of the present application. Without conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0041] Please refer to Figure 1 , a multi - band fusion image quality evaluation method provided by an embodiment of the present invention. The method specifically includes steps S101 - S104:

[0042] Step S101: Obtain the subjective evaluation scores and objective evaluation scores of multi - band fusion images respectively obtained by multiple fusion algorithms.

[0043] Among them, the fusion algorithms in this embodiment can adopt some existing fusion algorithms. Then, each fusion algorithm is used to fuse visible light images, long - wave infrared images, and short - wave infrared images to obtain multi - band fusion images. After obtaining the multi - band fusion images corresponding to each fusion, the multi - band fusion images are evaluated through subjective evaluation methods and objective evaluation methods to obtain subjective evaluation scores and objective evaluation scores. The subjective evaluation score is the evaluation of the multi - band fusion image by the subjective vision of the tester, and the objective evaluation score is the evaluation of the multi - band fusion image using existing image evaluation algorithms.

[0044] In an optional embodiment provided by the present application, the obtaining of the subjective evaluation scores and objective evaluation scores of multi - band fusion images respectively obtained by multiple fusion algorithms includes: fusing the sample images in the test dataset through the fusion algorithm to be evaluated and the benchmark fusion algorithm respectively to obtain the multi - band fusion image to be evaluated and the benchmark multi - band fusion image; determining the subjective evaluation score of each multi - band fusion image to be evaluated according to the evaluation score of the tester on the multi - band fusion image to be evaluated relative to the benchmark multi - band fusion image; evaluating the multi - band fusion image to be evaluated through multiple objective evaluation indicators to obtain the objective evaluation scores corresponding to the multiple objective evaluation indicators respectively. Among them, the fusion algorithm to be evaluated is the fusion algorithm that needs to be evaluated, and the benchmark fusion algorithm is an existing fusion algorithm for evaluating the fusion algorithm to be evaluated. The objective evaluation indicators can specifically be average gradient (AG), spatial frequency (SF), contrast (SD), structural similarity index (SSIM), information entropy (EN), etc. This embodiment does not make specific limitations on this.

[0045] It should be noted that since the difficulty of subjective visual evaluation lies in being affected by personal subjective factors, a scientific subjective evaluation procedure should be studied and formulated to reduce the subjective one - sidedness of the evaluation. Therefore, in this embodiment, the tester is required to compare the multi - band fusion image to be evaluated and the benchmark multi - band fusion image to determine the subjective evaluation score of the multi - band fusion image to be evaluated. This subjective evaluation method reduces the error caused by the personal subjective factors of the tester in the traditional subjective evaluation process, and at the same time, through the standardized evaluation process, ensures the fairness and credibility of the evaluation results.

[0046] Such as Figure 2The figure shows a flowchart of a method for obtaining subjective evaluation scores provided by an embodiment. This method uses a fusion algorithm to be evaluated and a benchmark fusion algorithm to fuse the sample data of multiple scenarios in a test dataset to obtain a multi-band fusion image to be evaluated and a benchmark multi-band fusion image. Then, the tester compares and evaluates the multi-band fusion image to be evaluated and the benchmark multi-band fusion image pairwise to obtain the number of votes for the multi-band fusion image to be evaluated. Finally, all the votes obtained for the multi-band fusion image to be evaluated are accumulated to obtain the subjective evaluation score of the fusion algorithm to be evaluated. As Figure 2 shown, for images captured by different imaging methods in a certain scenario i, the fusion algorithm to be evaluated and N benchmark fusion algorithms are respectively used to generate a multi-band fusion image to be evaluated (i0) and benchmark multi-band fusion images (i1,…,iN). Then, multiple different comparison combinations are formed by combining i0 with i1-iN, and the tester selects the image with better fusion effect according to subjective evaluation indicators.

[0047] Specifically, taking Figure 2 the subjective quality evaluation of images i0 and i1 as an example. If i0 is better, it is equivalent to adding 1 vote to both the "number of votes" and the "total number of votes"; if i0 is not better, there are two cases: if it is selected to skip this evaluation, it is equivalent to adding 0.5 vote to both the "number of votes" and the "total number of votes"; otherwise, it is equivalent to adding 1 vote to the "total number of votes" without increasing the "number of votes". Finally, if the test is completed, calculate the number of votes / total number of votes to obtain the final score (the subjective score of the algorithm to be evaluated); if the test is not completed, proceed to the next group of images for continued testing, and accumulate the number of votes and the total number of votes. Then, according to the voting situation of the tester, the subjective evaluation score of the fusion algorithm to be evaluated in this scenario is output according to the Figure 2 rules in. After all testers' visual subjective evaluations of all scenarios, the subjective evaluation score of the fusion algorithm to be evaluated is finally obtained through comprehensive calculation as shown in Table 1 (Table 1 is an example of the evaluation results of a single tester).

[0048] Table 1

[0049] Scenario 1 Scenario 2 ... Scenario n Average Score Algorithm 1 to be Evaluated 5.71 x ... x 5.25 ... ... ... ... ... ... Algorithm x to be Evaluated 8.62 x ... x 8.20

[0050] Step S102, calculate the correlation coefficient according to the subjective evaluation score and the objective evaluation score.

[0051] To evaluate the matching degree between each objective evaluation index and the subjective evaluation result, the Spearman correlation coefficient is adopted in this embodiment. The Spearman correlation coefficient is a non-parametric statistical method used to measure the monotonic correlation between two ranking sequences. The higher the value, the stronger the consistency. Among them, the monotonic correlation describes whether there is a consistent monotonic relationship between two variables. Specifically, if one variable monotonically increases as another variable increases, or monotonically decreases as the other variable decreases, then there is a monotonic correlation between the two variables, and the objective evaluation index can better reflect the actual fusion effect. In this embodiment, by performing a consistency test on the score vector of each objective index and the reference vector of the subjective evaluation result, the correlation coefficient of each objective index is obtained. The higher the correlation coefficient value, the more accurately the index can reflect the quality characteristics of the subjective evaluation, and thus the higher its applicability and importance.

[0052] As Figure 3 shown, it is a schematic diagram of a correlation coefficient calculation process provided by this embodiment. In this embodiment, first, the source images in the evaluation dataset are respectively used for the fusion of the fusion algorithm to be evaluated to obtain the multi-band fusion images to be evaluated, and the benchmark fusion algorithm is used for fusion to obtain multiple benchmark multi-band fusion images (that is, the 8 fusion images obtained by 8 benchmark fusion algorithms in Figure 3 ), then the evaluator conducts a visual subjective evaluation on the multi-band fusion images to be evaluated to obtain the subjective evaluation scores, conducts 28 objective index evaluations on the 8 benchmark multi-band fusion images respectively to obtain the objective evaluation scores of each objective evaluation index, and then calculates the correlation coefficient between the subjective evaluation scores and the objective evaluation. Among them, the objective evaluation indexes include average gradient (AG), spatial frequency (SF), contrast (SD), structural similarity index (SSIM), information entropy (EN), etc. In this embodiment, before calculating the correlation coefficient, it is also necessary to normalize the subjective evaluation scores and the objective evaluation scores to ensure the objectivity and fairness of the evaluation results.

[0053] For example, for the PFF fusion image, 28 objective evaluation score vectors are calculated, and then the correlation coefficients corresponding to the subjective evaluation score vector of the multi-band fusion image to be evaluated and the 28 objective evaluation score vectors in the PFF fusion image are calculated respectively.

[0054] Step S103, determine the positive objective evaluation index corresponding to each evaluation dimension from the objective evaluation indexes for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient.

[0055] Specifically, from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient, the positive objective evaluation indicators corresponding to each evaluation dimension are determined, including: among the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions, the top N objective evaluation indicators of the correlation coefficient are selected as the positive objective evaluation indicators for the corresponding evaluation dimension. In this embodiment, the objective evaluation indicators with the top three correlation coefficients (if the number of dimensions is less than three, all indicators are selected) in the evaluation dimension can be screened out as the positive objective evaluation indicators.

[0056] Step S104, calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension.

[0057] In this embodiment, 6 evaluation dimensions can be defined for the multi-band fusion image: overall visual perception and visual saliency, richness of picture details, information volume of the fusion picture, edge strength of objects (such as road surfaces, signs, etc.), similarity to the source image, recognizability and cognition of visual references, etc. This embodiment does not make specific limitations. The image quality score of the multi-band fusion image is determined jointly by multiple evaluation dimensions.

[0058] In an optional embodiment provided by the present application, the calculating the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension includes:

[0059] Step S1041, obtain the Spearman rank correlation coefficient of the positive objective evaluation indicators corresponding to each evaluation dimension, and calculate the comprehensive objective evaluation score according to the Spearman rank correlation coefficient of the positive objective evaluation indicators corresponding to each evaluation dimension.

[0060] Specifically, the calculating the comprehensive objective evaluation score according to the Spearman rank correlation coefficient of the positive objective evaluation indicators corresponding to each evaluation dimension includes: obtaining the weight value of the positive objective evaluation indicators of each evaluation dimension; performing weighted calculation on the weight values and positive objective evaluation indicators of the positive objective evaluation indicators of all evaluation dimensions to obtain the comprehensive objective evaluation score of each evaluation dimension.

[0061] It should be noted that different evaluation dimensions are evaluated using corresponding objective evaluation indicators. For example, Figure 4 the richness of picture details is evaluated using three objective evaluation indicators, namely AG, SF, and SD. In this embodiment, if the number of objective evaluation indicators of an evaluation dimension exceeds 3, the top 3 objective evaluation indicators of the correlation coefficient are taken as the positive objective evaluation indicators of this evaluation dimension; if the number of objective evaluation indicators of an evaluation dimension is less than or equal to 3, the corresponding objective evaluation indicators are directly taken as the positive objective evaluation indicators. For example, Figure 4For the object edge strength, picture detail richness, and recognition degree of the target reference object class schedule in it, the corresponding objective evaluation indicators are directly used as positive objective evaluation indicators.

[0062] Determine the initial weight value of the positive objective evaluation indicator under the corresponding evaluation dimension according to the ratio of the correlation coefficient of the positive objective evaluation indicator to the total score of the correlation coefficients of all positive objective evaluation indicators under the evaluation dimension. Further, in this embodiment, the particle swarm filtering algorithm is used to update the weight value of the positive objective evaluation indicator.

[0063] Specifically, each evaluation indicator has a score vector, and the score vector of each indicator is Y i =[y i1 ,y i2 ,...,y im , where i = 1, 2,..., n, m is the length of each score vector, and the reference vector R = [r1, r2,..., r m , and this reference vector is the target that the comprehensive score vector is expected to approach. The comprehensive score vector Y W is obtained by weighted summation of all indicator score vectors:

[0064]

[0065] Among them, w i is the weight coefficient of each positive objective evaluation indicator. The weight coefficients w i , w2,..., w n are optimized by the particle swarm optimization algorithm (PSO) to maximize the correlation coefficient between the comprehensive score vector Y W and the reference vector R.

[0066] Step S1042, perform weighted calculation on the subjective evaluation score and the comprehensive objective evaluation score to obtain the image quality score of the multi-band fusion image.

[0067] According to the experience and ability levels of different evaluators, the present invention sets the adjustment rules for the subjective and objective weights. Specifically, the following table lists the corresponding subjective and objective weight distributions for different evaluator groups (such as high-level evaluators, medium-level pilots, and junior pilots), where high-level evaluators are experienced pilots, medium-level evaluators are pilots with certain experience but not reaching the expert level, and low-level evaluators are junior pilots with less experience, showing the relationship between the experience level of the evaluators and the subjective and objective weights.

[0068] Table 2

[0069] Evaluators Subjective Proportion Objective Proportion High-level Evaluators 0.7 0.3 Medium-level Evaluators 0.5 0.5 Low-level Evaluators 0.3 0.7

[0070] Through this principle, pilots with different experience levels will adjust the weights of subjective and objective scores according to their experience level, so that the evaluation results can better meet the actual application requirements.

[0071] As Figure 5 shown, a multi-band fusion image quality evaluation principle diagram provided in this embodiment is a comprehensive scoring method that combines objective evaluation and subjective evaluation. By optimizing the index weights and scoring models, the scientificity and reliability of image fusion quality evaluation are improved. The weighted combination of subjective and objective scores based on the pilot's experience level provides a flexible evaluation method suitable for different evaluators and application scenarios. In this embodiment, objective evaluation indicators with higher importance in each dimension are first screened out, and the weights of these indicators are further optimized through the particle swarm optimization algorithm. Then, the subjective score is determined according to the subjective evaluation process and method. Finally, the optimized subjective and objective scores are weighted and summed according to reasonable weights to obtain the final comprehensive evaluation score of the image quality. The comprehensive evaluation system of this embodiment mainly includes three important parts,

[0072] (1) Subjective evaluation: Combining the actual perception evaluation of pilots or experts to provide scores that conform to human eye perception. The pilot scores the image according to the perception characteristics such as image clarity, detail retention, and color contrast to obtain the subjective evaluation score.

[0073] (2) Objective evaluation: Based on statistical methods, the objective evaluation indicators that are highly consistent with the subjective evaluation are screened to quantify the image fusion quality and obtain the objective evaluation score.

[0074] (3) Comprehensive evaluation: Combining the optimized subjective and objective weight coefficients, the subjective score and the objective score are weighted and summed to obtain the final comprehensive evaluation score. In order to adapt to evaluators with different experience levels, the present invention sets the weight adjustment rules for subjective and objective scores. Specifically, pilots with different experience levels will adjust the weights of subjective and objective scores according to their experience level. The final comprehensive evaluation score is obtained by weighted summation of the subjective score and the objective score.

[0075] An embodiment of the present invention provides a multi-band fusion image quality evaluation method. First, obtain the subjective evaluation scores and objective evaluation scores of multi-band fusion images obtained by various fusion algorithms respectively. Then, calculate the correlation coefficient according to the subjective evaluation scores and the objective evaluation scores. After that, determine the positive objective evaluation index corresponding to each evaluation dimension from the objective evaluation indexes for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient. Finally, calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation index corresponding to each evaluation dimension. Since the image quality score of the multi-band fusion image in this application comprehensively considers the subjective evaluation score and the positive objective evaluation index, and this positive objective evaluation index can represent the consistency between the objective index and the subjective evaluation, the image quality score of the multi-band fusion image obtained through this application can reflect the visual perception effect of the pilot during actual flight. Furthermore, the accuracy of multi-band fusion image quality evaluation can be improved through this application.

[0076] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0077] In one embodiment, a multi-band fusion image quality evaluation device is provided, and this multi-band fusion image quality evaluation device corresponds one-to-one with the multi-band fusion image quality evaluation method in the above embodiment. As Figure 6 shown, the detailed description of each functional module of the multi-band fusion image quality evaluation device is as follows:

[0078] An acquisition module 61, configured to acquire the subjective evaluation scores and objective evaluation scores of multi-band fusion images obtained by various fusion algorithms respectively;

[0079] A calculation module 62, configured to calculate the correlation coefficient according to the subjective evaluation scores and the objective evaluation scores;

[0080] A determination module 63, configured to determine the positive objective evaluation index corresponding to each evaluation dimension from the objective evaluation indexes for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient;

[0081] The calculation module 62 is further configured to calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation index corresponding to each evaluation dimension.

[0082] In an alternative embodiment, the acquisition module 61 is specifically configured to:

[0083] The sample images in the test dataset are respectively fused by the fusion algorithm to be evaluated and the benchmark fusion algorithm to obtain the multi-band fusion images to be evaluated and the benchmark multi-band fusion images;

[0084] According to the evaluation scores of the tester on the multi-band fusion image to be evaluated relative to the benchmark multi-band fusion image, determine the subjective evaluation scores of each multi-band fusion image to be evaluated;

[0085] Evaluate the multi-band fusion images to be evaluated through a variety of objective evaluation indicators to obtain the objective evaluation scores corresponding to the various objective evaluation indicators respectively.

[0086] In an alternative embodiment, the determination module 63 is specifically configured to:

[0087] Among the objective evaluation indicators for the evaluation EVS application scenario under different evaluation dimensions, select the first N objective evaluation indicators with the highest correlation coefficients as the positive objective evaluation indicators for the corresponding evaluation dimensions.

[0088] In an alternative embodiment, the calculation module 62 is specifically configured to:

[0089] Obtain the Spearman rank correlation coefficients of the positive objective evaluation indicators corresponding to each evaluation dimension, and calculate the comprehensive objective evaluation score according to the Spearman rank correlation coefficients of the positive objective evaluation indicators corresponding to each evaluation dimension;

[0090] Perform weighted calculation on the subjective evaluation score and the comprehensive objective evaluation score to obtain the image quality score of the multi-band fusion image.

[0091] In an alternative embodiment, the calculation module 62 is specifically configured to:

[0092] Obtain the weight values of the positive objective evaluation indicators for each evaluation dimension;

[0093] Perform weighted calculation on the weight values and the positive objective evaluation indicators of the positive objective evaluation indicators for all evaluation dimensions to obtain the comprehensive objective evaluation score for each evaluation dimension.

[0094] In an alternative embodiment, the calculation module 62 is further configured to:

[0095] Determine the initial weight value of the positive objective evaluation indicator under the corresponding evaluation dimension according to the ratio of the correlation coefficient of the positive objective evaluation indicator to the total score of the correlation coefficients of all positive objective evaluation indicators under the evaluation dimension.

[0096] In an alternative embodiment, the calculation module 62 is further configured to:

[0097] Update the weight value of the positive objective evaluation indicator by using the particle swarm filtering algorithm.

[0098] For the specific limitations of the multi - band fusion image quality evaluation device, reference can be made to the limitations of the multi - band fusion image quality evaluation method in the above text, which will not be elaborated here. Each module in the above device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0099] In one embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 7 shown. The computer device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non - volatile storage medium and an internal memory. The non - volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non - volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a multi - band fusion image quality evaluation method.

[0100] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0101] Obtain the subjective evaluation scores and objective evaluation scores of the multi - band fusion images obtained by different fusion algorithms respectively;

[0102] Calculate the correlation coefficient according to the subjective evaluation score and the objective evaluation score;

[0103] Determine the positive objective evaluation index corresponding to each evaluation dimension from the objective evaluation indexes for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient;

[0104] Calculate the image quality score of the multi - band fusion image according to the subjective evaluation score and the positive objective evaluation index corresponding to each evaluation dimension.

[0105] In one embodiment, a computer - readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the following steps are implemented:

[0106] Obtain the subjective evaluation scores and objective evaluation scores of the multi - band fusion images obtained by different fusion algorithms respectively;

[0107] Calculate the correlation coefficient based on the subjective evaluation score and the objective evaluation score;

[0108] Determine the positive objective evaluation indicators corresponding to each evaluation dimension from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient;

[0109] Calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension.

[0110] In one embodiment, a computer program product is provided. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0111] Obtain the subjective evaluation score and the objective evaluation score of the multi-band fusion image obtained by multiple fusion algorithms respectively;

[0112] Calculate the correlation coefficient based on the subjective evaluation score and the objective evaluation score;

[0113] Determine the positive objective evaluation indicators corresponding to each evaluation dimension from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient;

[0114] Calculate the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension.

[0115] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer 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 in the present application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory 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 DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0116] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0117] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. 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 described in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention and should all be included in the protection scope of the present invention.

Claims

1. A multi-band fusion image quality evaluation method, characterized in that The method includes: Obtaining the subjective evaluation scores and objective evaluation scores of multi-band fusion images respectively obtained by multiple fusion algorithms; Calculating a correlation coefficient according to the subjective evaluation score and the objective evaluation score; Determining positive objective evaluation indicators corresponding to each evaluation dimension from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient; Calculating the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension.

2. The method according to claim 1, characterized in that, The obtaining the subjective evaluation scores and objective evaluation scores of multi-band fusion images respectively obtained by multiple fusion algorithms includes: Performing fusion on the sample images in the test dataset through the fusion algorithm to be evaluated and the benchmark fusion algorithm respectively to obtain the multi-band fusion image to be evaluated and the benchmark multi-band fusion image; Determining the subjective evaluation score of each multi-band fusion image to be evaluated according to the evaluation score of the multi-band fusion image to be evaluated relative to the benchmark multi-band fusion image by the tester; Evaluating the multi-band fusion image to be evaluated through multiple objective evaluation indicators to obtain objective evaluation scores corresponding to the multiple objective evaluation indicators respectively.

3. The method according to claim 1, characterized in that The determining positive objective evaluation indicators corresponding to each evaluation dimension from the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions through the correlation coefficient includes: Among the objective evaluation indicators for evaluating the EVS application scenario under different evaluation dimensions, selecting the first N objective evaluation indicators of the correlation coefficient as the positive objective evaluation indicators for the corresponding evaluation dimension.

4. The method according to claim 3, characterized in that, The calculating the image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation indicators corresponding to each evaluation dimension includes: Obtaining the correlation coefficient of the positive objective evaluation indicator corresponding to each evaluation dimension, and calculating a comprehensive objective evaluation score according to the correlation coefficient of the positive objective evaluation indicator corresponding to each evaluation dimension; Performing weighted calculation on the subjective evaluation score and the comprehensive objective evaluation score to obtain the image quality score of the multi-band fusion image.

5. The method according to claim 4, wherein The calculating the comprehensive objective evaluation score according to the correlation coefficient of the positive objective evaluation indicator corresponding to each evaluation dimension includes: Obtaining the weight value of the positive objective evaluation indicator of each evaluation dimension; Performing weighted calculation on the weight values and positive objective evaluation indicators of the positive objective evaluation indicators of all evaluation dimensions to obtain the comprehensive objective evaluation score of each evaluation dimension.

6. The method according to claim 5, characterized in that, The method further includes: Determining the initial weight value of the positive objective evaluation indicator in the corresponding evaluation dimension according to the ratio of the correlation coefficient of the positive objective evaluation indicator to the total score of the correlation coefficients of all positive objective evaluation indicators in the evaluation dimension; 7. The method according to claim 6, characterized in that, The method further includes: Updating the weight value of the positive objective evaluation indicator by using a particle swarm filtering algorithm.

8. A multi-band fusion image quality evaluation device, characterized in that The device includes: An obtaining module, configured to obtain the subjective evaluation scores and objective evaluation scores of multi-band fusion images respectively obtained by multiple fusion algorithms; A calculating module, configured to calculate a correlation coefficient according to the subjective evaluation score and the objective evaluation score; A determination module, configured to determine, from the objective evaluation indexes for evaluating the EVS application scenario under different evaluation dimensions by using the correlation coefficient, the positive objective evaluation index corresponding to each evaluation dimension respectively; The calculation module is further configured to calculate an image quality score of the multi-band fusion image according to the subjective evaluation score and the positive objective evaluation index corresponding to each evaluation dimension respectively.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the multi-band fusion image quality evaluation method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the multi-band fusion image quality evaluation method according to any one of claims 1 to 7 is implemented.

Images