Remote sensing image recognition method and device based on multi-index fusion

Abstract

The embodiment of the invention discloses a remote sensing image recognition method and device based on multi-index fusion. A specific embodiment of the method comprises the following steps: intercepting each remote sensing image in a remote sensing image sequence; determining a side lobe intensity value and a main lobe intensity value corresponding to a point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence to obtain a main lobe intensity value sequence and a side lobe intensity value sequence; performing transformation processing on each intercepted remote sensing image in the intercepted remote sensing image sequence to generate a complex number self-correlation feature value sequence; performing multi-dimensional index detection on each intercepted remote sensing image in the intercepted remote sensing image sequence to generate an index detection result; performing comprehensive quality detection on each remote sensing image in the remote sensing image sequence; and performing image recognition on each remote sensing image in the remote sensing image sequence to generate an image recognition result. According to the embodiment, the accuracy of a remote sensing image quality detection result can be improved.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a remote sensing image recognition method and device based on multi-index fusion. Background Art

[0002] Remote sensing image recognition is a technology used to detect remote sensing image recognition. At present, the method commonly used in remote sensing image recognition is to obtain high-resolution radar images of the surface by emitting radar pulses and recording the returned echoes. Unlike optical remote sensing, SAR (Synthetic Aperture Radar) is not affected by weather conditions, light and clouds, so it is suitable for all-weather and all-day monitoring and surface observation. SAR image quality evaluation plays an important role in indicating the practical application of SAR images, and is of great significance for improving the level of subsequent target detection, segmentation, fusion and other applications.

[0003] However, in practice, it is found that when the above method is used for remote sensing image recognition, the following technical problems often occur:

[0004] First, the selection of evaluation indicators is not comprehensive enough. Only multiple objective indicators in the area target evaluation indicators are selected, which makes it difficult to comprehensively evaluate the image quality from all aspects, thus reducing the accuracy of remote sensing image quality detection results, and further reducing the accuracy of image recognition;

[0005] Second, it is difficult to comprehensively process multi-dimensional indicators, and using each single indicator to describe image quality is not comprehensive enough. At the same time, since the factors affecting quality detection are interfered by various factors, the setting of the optimal weight (that is, the optimal weight selected by the test experiment) is likely to bring bias to quality detection, resulting in deviations in detection results. Therefore, it is difficult to efficiently combine multiple indicators for comprehensive evaluation, which leads to reduced efficiency and accuracy of quality detection.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0007] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0008] Some embodiments of the present disclosure propose a remote sensing image quality detection method and device based on multi-index fusion to solve one or more of the technical problems mentioned in the above background technology section.

[0009] In a first aspect, some embodiments of the present disclosure provide a remote sensing image recognition method based on multi-index fusion, the method comprising: acquiring a remote sensing image sequence; performing interception processing on each remote sensing image in the remote sensing image sequence to generate a sequence of intercepted remote sensing images, wherein each intercepted remote sensing image in the intercepted remote sensing image sequence includes a point target; determining the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence to obtain a main lobe intensity value sequence and a side lobe intensity value sequence; performing transformation processing on each intercepted remote sensing image in the intercepted remote sensing image sequence to generate a sequence of complex autocorrelation eigenvalues; based on The above-mentioned main lobe intensity value sequence and the above-mentioned side lobe intensity value sequence are used to perform multi-dimensional index detection on each of the above-mentioned intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence to generate an index detection result, wherein the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category; based on the above-mentioned complex autocorrelation eigenvalue sequence and the above-mentioned index detection results, a comprehensive quality detection is performed on each of the remote sensing images in the remote sensing image sequence to generate an image quality detection value sequence; based on the above-mentioned image quality detection value sequence, image recognition is performed on each of the remote sensing images in the above-mentioned remote sensing image sequence to generate an image recognition result.

[0010] In a second aspect, some embodiments of the present disclosure provide a remote sensing image recognition device based on multi-index fusion, the device comprising: an acquisition unit, configured to acquire a remote sensing image sequence; a capture processing unit, configured to capture and process each remote sensing image in the above remote sensing image sequence to generate a captured remote sensing image sequence, wherein each captured remote sensing image in the above captured remote sensing image sequence includes a point target; a determination unit, configured to determine the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each captured remote sensing image in the above captured remote sensing image sequence, and obtain a main lobe intensity value sequence and a side lobe intensity value sequence; a transformation processing unit, configured to transform each captured remote sensing image in the above captured remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence; The multi-dimensional index detection unit is configured to perform multi-dimensional index detection on each remote sensing image in the above-mentioned intercepted remote sensing image sequence based on the above-mentioned main lobe intensity value sequence and the above-mentioned side lobe intensity value sequence to generate an index detection result, wherein the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category; the comprehensive quality detection unit is configured to perform comprehensive quality detection on each remote sensing image in the remote sensing image sequence based on the above-mentioned complex autocorrelation eigenvalue sequence and the above-mentioned index detection result to generate an image quality detection value sequence; the image recognition unit is configured to perform image recognition on each remote sensing image in the above-mentioned remote sensing image sequence based on the above-mentioned image quality detection value sequence to generate an image recognition result.

[0011] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the above-mentioned first aspect.

[0012] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.

[0013] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the remote sensing image recognition method based on multi-index fusion of some embodiments of the present disclosure, the accuracy of remote sensing image recognition can be improved. Specifically, the reason for the reduction in the accuracy of remote sensing image recognition is that the selection of evaluation indicators is not comprehensive enough. Only multiple objective indicators in the surface target evaluation indicators are selected, and it is difficult to comprehensively evaluate the image quality from various aspects. Based on this, the remote sensing image recognition method based on multi-index fusion of some embodiments of the present disclosure first obtains a remote sensing image sequence. Then, each remote sensing image in the above remote sensing image sequence is intercepted to generate a sequence of intercepted remote sensing images. Among them, each intercepted remote sensing image in the above intercepted remote sensing image sequence includes a point target. By intercepting the area where the point target is located, it is convenient to compare the quality of remote sensing images of consecutive frames. Then, the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the above intercepted remote sensing image sequence are determined to obtain a main lobe intensity value sequence and a side lobe intensity value sequence. By introducing the sidelobe intensity value and the mainlobe intensity value, they can be used as the basic data of the point target evaluation index to participate in the index calculation. After that, each of the intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence is transformed to generate a complex autocorrelation feature value sequence. Here, the transformation process can be used to determine the coherence of the image area from the complex level of the pixels in the image, thereby reflecting the evaluation degree of the image in terms of coherence characteristics. Then, based on the above-mentioned mainlobe intensity value sequence and the above-mentioned sidelobe intensity value sequence, each of the intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence is subjected to multi-dimensional index detection to generate an index detection result. Among them, the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category. Here, through multi-dimensional index detection, index detection can be performed from at least the following dimensions: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category, thereby realizing multi-dimensional evaluation. In this way, the comprehensiveness of the index can be improved so that quality detection can be objectively performed from multiple aspects. Finally, based on the above complex autocorrelation eigenvalue sequence and the above index detection results, a comprehensive quality detection is performed on each remote sensing image in the remote sensing image sequence to generate an image quality detection value sequence. Based on the above image quality detection value sequence, image recognition is performed on each remote sensing image in the above remote sensing image sequence to generate an image recognition result. Here, through comprehensive quality detection, multiple dimensional indicators can be used for comprehensive quality evaluation. Therefore, more comprehensive multi-dimensional indicators can be applied to image quality detection to improve the accuracy of remote sensing image quality detection results. In turn, the accuracy of image recognition is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0015] Figure 1 is a flowchart of some embodiments of the remote sensing image recognition method based on multi-index fusion according to the present disclosure;

[0016] Figure 2 It is a structural schematic diagram of some embodiments of the remote sensing image recognition device based on multi-index fusion according to the present disclosure;

[0017] Figure 3 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0018] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0019] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0020] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0021] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0022] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0023] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0024] Figure 1The process 100 of some embodiments of the remote sensing image recognition method based on multi-index fusion according to the present disclosure is shown. The remote sensing image recognition method based on multi-index fusion includes the following steps:

[0025] Step 101: Acquire a remote sensing image sequence.

[0026] In some embodiments, the execution subject of the remote sensing image recognition method based on multi-index fusion can obtain the remote sensing image sequence by wired or wireless means, wherein each remote sensing image in the remote sensing image sequence can be a continuous frame image taken by a satellite.

[0027] It should be noted that the above-mentioned wireless connection methods may include but are not limited to 3G / 4G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future.

[0028] Step 102, intercepting each remote sensing image in the remote sensing image sequence to generate an intercepted remote sensing image sequence.

[0029] In some embodiments, the execution subject may perform interception processing on each remote sensing image in the remote sensing image sequence to generate an intercepted remote sensing image sequence. Each intercepted remote sensing image in the intercepted remote sensing image sequence includes a point target. Here, the point target may be the coordinates of a pre-complemented corner reflector.

[0030] In some optional implementations of some embodiments, the execution subject intercepts each remote sensing image in the remote sensing image sequence to generate a intercepted remote sensing image sequence, which may include the following steps:

[0031] For each remote sensing image in the above remote sensing image sequence, perform the following steps:

[0032] The first step is to project the preset point target coordinates into the remote sensing image to obtain a projected point target coordinate set. The point target coordinates can be projected from the world coordinate system into the coordinate system of the remote sensing image to obtain a projected point target coordinate set.

[0033] The second step is to determine the image clarity of each projected point target coordinate position in the projected point target coordinate set to obtain an image clarity set. The image clarity of each projected point target coordinate position in the projected point target coordinate set can be determined by a preset clarity algorithm to obtain an image clarity set.

[0034] As an example, the clarity algorithm may include but is not limited to at least one of the following: Laplace operator method, variance method, energy gradient method, etc.

[0035] The third step is to select the corresponding projected point target coordinates with the highest image clarity from the above projected point target coordinate set as the target point position coordinates.

[0036] The fourth step is to capture the image according to the preset capture boundary with the above target point position coordinates as the center to obtain the captured remote sensing image. The size of the capture boundary can be pre-selected.

[0037] Step 103, determining the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence, and obtaining a main lobe intensity value sequence and a side lobe intensity value sequence.

[0038] In some embodiments, the execution subject may determine the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence in various ways to obtain a main lobe intensity value sequence and a side lobe intensity value sequence, wherein each side lobe intensity value and main lobe intensity value corresponds to a intercepted remote sensing image.

[0039] In some optional implementations of some embodiments, the execution subject determines the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence, and obtains the main lobe intensity value sequence and the side lobe intensity value sequence, which may include the following steps:

[0040] For each intercepted remote sensing image in the above intercepted remote sensing image sequence, perform the following steps:

[0041] The first step is to obtain the impulse response function wave of the coordinate position of the target point in the above-mentioned intercepted remote sensing image, wherein the impulse response function wave includes a main lobe wave and multiple side lobe waves.

[0042] In the second step, the peak value of the impulse response function wave including the main lobe wave is determined as the main lobe intensity value.

[0043] In the third step, the peak value of the largest sidelobe wave among the multiple sidelobe waves included in the impulse response function wave is determined as the sidelobe intensity value.

[0044] Step 104, transforming each of the intercepted remote sensing images in the intercepted remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence.

[0045] In some embodiments, the execution entity may transform each of the intercepted remote sensing images in the intercepted remote sensing image sequence in various ways to generate a complex autocorrelation eigenvalue sequence.

[0046] In some optional implementations of some embodiments, the execution subject transforms each of the intercepted remote sensing images in the intercepted remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence, which may include the following steps:

[0047] For each intercepted remote sensing image in the above intercepted remote sensing image sequence, perform the following steps:

[0048] The complex values ​​of each pixel in the above-mentioned intercepted remote sensing image are transformed in two dimensions to generate complex autocorrelation eigenvalues. The complex values ​​of each pixel in the above-mentioned intercepted remote sensing image are transformed in two dimensions to generate complex autocorrelation eigenvalues ​​by the following formula:

[0049] CCF=F -1 (|F(I(x,y))| 2 ).

[0050] Where I(x, y) represents the complex value of the intercepted remote sensing image at the coordinate (x, y). 1 () denotes two-dimensional Fourier transform and two-dimensional inverse Fourier transform respectively. CCF denotes complex autocorrelation eigenvalue.

[0051] Step 105 , based on the main lobe intensity value sequence and the side lobe intensity value sequence, perform multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence to generate an index detection result.

[0052] In some embodiments, the execution subject may perform multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence in various ways to generate an index detection result. The index detection result may include but is not limited to at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category. Here, the point target evaluation category may include a quality assessment index of a point target in the image map, which is used to perform point target quality assessment. The surface target evaluation index category may include a regional quality assessment index in the image map, which is used to perform regional quality assessment. The information entropy evaluation index category may include an information entropy evaluation index value. Indicators for quality detection from the perspective of coordinate probability. The contrast evaluation index category may include a contrast evaluation index for quality assessment from the perspective of overall image contrast.

[0053] In some optional implementations of some embodiments, the execution subject performs multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result, which may include the following steps:

[0054] For each intercepted remote sensing image in the above intercepted remote sensing image sequence, perform the following steps:

[0055] The first step is to generate the image peak sidelobe ratio using the main lobe intensity value and sidelobe intensity value corresponding to the above-mentioned intercepted remote sensing image. The image peak sidelobe ratio can be generated by the following formula:

[0056]

[0057] Among them, I s is the maximum sidelobe intensity value among all sidelobe intensity values. m is the strength of the main lobe. PSLR stands for image peak sidelobe ratio.

[0058] In practice, the peak sidelobe ratio (PSLR) is defined as the ratio of the maximum sidelobe intensity to the main lobe intensity. This parameter can be used to measure the ability of SAR images to detect weak targets from around strong targets. The smaller the PSLR, the more concentrated the energy of the point target, and the more obvious the target is in the SAR image.

[0059] The second step is to determine the main lobe interval of the above-mentioned intercepted remote sensing image. The main lobe interval can be determined as the abscissa interval range of the main lobe in the impulse response function wave.

[0060] The third step is to determine the image integral sidelobe ratio of the above-mentioned intercepted remote sensing image within the above-mentioned main lobe interval. The image integral sidelobe ratio of the above-mentioned intercepted remote sensing image can be determined by the following formula:

[0061]

[0062] Wherein, h(τ) represents the impulse response function in the range or azimuth direction. [a, b] represents the main lobe interval. ISLR represents the image integrated sidelobe ratio.

[0063] In practice, the integrated sidelobe ratio is defined as the ratio of the sum of all sidelobe energies of the impulse response function to the mainlobe energy. This indicator can focus on describing the importance of the sidelobe relative to the mainlobe, that is, the ability to detect dark targets near bright targets in SAR images. In addition, when the energy of the target is dispersed, the peak sidelobe ratio and the integrated sidelobe will be relatively high. In the image, the strong target will appear in a cross shape to cover the surrounding dark targets. When the image contrast is low, the image is full of a large number of shadow areas. At this time, the value of ISLR is also relatively large. Generally, an excessively large ISLR will cause a large number of false targets to appear during the target detection process. Therefore, by introducing the image integrated sidelobe ratio, it can be used for more comprehensive image quality detection, thereby improving the accuracy of quality detection.

[0064] The fourth step is to determine the ratio of the number of image pixels corresponding to the impulse response function wave corresponding to the above-mentioned intercepted remote sensing image at the preset decibel position to the pixel resolution, and determine the average of each ratio as the spatial resolution index value. There are multiple pixel positions in the intercepted remote sensing image corresponding to the preset decibel value (for example, 3 decibel) of the impulse response function wave.

[0065] In the fifth step, the image peak sidelobe ratio, the image integral sidelobe ratio and the spatial resolution index value are determined as the index detection results representing the point target evaluation index category.

[0066] Optionally, the execution subject performs multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result, and may further include the following steps:

[0067] The first step is to grayscale each of the intercepted remote sensing images in the above intercepted remote sensing image sequence to obtain a grayscale remote sensing image sequence. The intercepted remote sensing images can be grayscaled to serve as grayscale remote sensing images.

[0068] In the second step, for each intercepted remote sensing image in the above intercepted remote sensing image sequence, the following steps are performed:

[0069] Step 1, using the length value and width value of the grayscale remote sensing image corresponding to the above-mentioned intercepted remote sensing image, determine the pixel intensity average value, pixel intensity variance value and pixel intensity standard deviation of the above-mentioned intercepted remote sensing image. Among them, first, the average value of the sum of the pixel grayscale values ​​of each coordinate position in the above-mentioned intercepted remote sensing image can be determined as the pixel intensity average value. Then, the pixel intensity variance value and pixel intensity standard deviation of the intercepted remote sensing image can be determined by using the variance formula and the standard deviation formula using the pixel intensity average value.

[0070] Step 2: Determine the equivalent viewing angle of the above-cut remote sensing image by taking the square value of the ratio of the above-mentioned pixel intensity average value to the above-mentioned pixel intensity standard deviation.

[0071] Step 3: Determine the radiometric resolution of the intercepted remote sensing image using the pixel intensity average, the pixel intensity standard deviation and the equivalent visual number. The radiometric resolution can be generated by the following formula:

[0072]

[0073] Among them, ENL represents the equivalent number of views. σ is the standard deviation of pixel intensity. μ represents the average pixel intensity. γ represents the radiometric resolution.

[0074] In the fourth step, the above pixel intensity average value, the above pixel intensity variance value, the above equivalent visual number and the above radiation resolution are determined as the indicator detection results representing the evaluation category of the surface target.

[0075] Optionally, the execution subject performs multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result, and may further include the following steps:

[0076] The first step is to determine the information entropy evaluation index value of each intercepted remote sensing image in the above intercepted remote sensing image sequence. The information entropy evaluation index value can be generated by the following formula:

[0077]

[0078]

[0079] Where H(f) is the information entropy evaluation index value. I(m, n) represents the grayscale value of the pixel at the pixel position (m, n) in the grayscale remote sensing image corresponding to the intercepted remote sensing image. M represents the maximum horizontal coordinate value of the intercepted remote sensing image. N represents the maximum vertical coordinate value of the intercepted remote sensing image. mn is the probability that I(m,n) is in the remote sensing image after interception.

[0080] In practice, the smaller the information entropy evaluation index value is, the more information the image contains and the better its focusing performance is. Conversely, the worse the image focusing is, the less clear the image is.

[0081] The second step is to determine the contrast evaluation index of each intercepted remote sensing image in the intercepted remote sensing image sequence based on the grayscale remote sensing image sequence. The contrast evaluation index can be generated by the following formula:

[0082]

[0083] Where, E(·) represents the expectation. C represents the contrast evaluation index.

[0084] In the third step, the above-mentioned information entropy evaluation index value and the above-mentioned contrast evaluation index are determined as the index detection result.

[0085] In practice, image contrast reflects the change in pixel intensity. The greater the contrast, the clearer the image intensity, which means the greater the probability of the existence of strong scattering points. Conversely, the possibility of the existence of strong scattering points is smaller.

[0086] Step 106 , based on the complex autocorrelation eigenvalue sequence and the index detection result, comprehensive quality detection is performed on each remote sensing image in the remote sensing image sequence to generate an image quality detection value sequence.

[0087] In some embodiments, the execution entity may perform comprehensive quality detection on each remote sensing image in the remote sensing image sequence based on the complex autocorrelation eigenvalue sequence and the index detection result to generate an image quality detection value sequence.

[0088] In some optional implementations of some embodiments, the execution subject performs comprehensive quality detection on each remote sensing image in the remote sensing image sequence based on the complex autocorrelation eigenvalue sequence and the index detection result to generate an image quality detection value sequence, which may include the following steps:

[0089] The first step is to use the above complex autocorrelation eigenvalue sequence and the above index detection results to generate a comprehensive index value sequence corresponding to each remote sensing image. Among them, the initial weight group can be obtained first. Here, the initial weight group does not need to be selected based on experiments, but can be divided according to the degree of influence of each index on quality detection, and the sum is 1. Secondly, the comprehensive index value sequence can be generated by the following formula:

[0090]

[0091] Where S represents the comprehensive index value in the comprehensive index value sequence. Each comprehensive index value corresponds to an index in the index detection result. W represents the initial weight. The initial weight corresponds to the index in the index detection result one by one. Θ represents the synthesis operator of the fuzzy matrix. ij It represents the index value of a certain evaluation index when a certain evaluation result is obtained during evaluation. If the index detection result corresponds to an evaluation category (for example, point target evaluation index category, surface target evaluation index category) including K evaluation factors, then a K×j fuzzy matrix will be generated, and its corresponding weight coefficient is (W i1 , W i2 …W ik ). Then r ij The value model of can be: ij =(W i1 , W i2 , …, W iK )·r Kj(u) . u represents a certain evaluation index.

[0092] In the second step, the comprehensive index value sequence is used to perform quality assessment on each remote sensing image in the remote sensing image sequence to generate an image quality detection value sequence. The comprehensive index value sequence can be determined as the image quality detection value sequence.

[0093] The above formulas and their related contents, as an inventive point of the embodiments of the present disclosure, solve the second technical problem mentioned in the background technology: "It is difficult to comprehensively process multi-dimensional indicators, and the use of each single indicator to describe the image quality is not comprehensive enough. At the same time, since the factors affecting quality detection are interfered by various factors, the setting of the optimal weight (that is, the optimal weight selected by the test experiment) is easy to bring bias to the quality detection, resulting in deviations in the detection results. Therefore, it is difficult to efficiently combine multiple indicators for comprehensive evaluation, thereby reducing the efficiency and accuracy of quality detection." The factors that reduce the efficiency and accuracy of quality detection are often as follows: it is difficult to comprehensively process multi-dimensional indicators, and the use of each single indicator to describe the image quality is not comprehensive enough. At the same time, since the factors affecting quality detection are interfered by various factors, the setting of the optimal weight (that is, the optimal weight selected by the test experiment) is easy to bring bias to the quality detection, resulting in deviations in the detection results. Therefore, it is difficult to efficiently combine multiple indicators for comprehensive evaluation. In order to achieve the effect of improving the efficiency and accuracy of quality detection, firstly, the corresponding indicator values ​​are generated through the above formulas. Then, considering that a single indicator of each indicator is not comprehensive enough to describe the image quality, each indicator can be processed comprehensively by generating a formula for the comprehensive indicator value. At the same time, considering the optimal weights set in advance through experiments, when the image is disturbed to varying degrees by various factors, it is easy to cause deviations, making the comprehensive indicators tend to be biased towards certain indicators. Therefore, by setting a more stable initial weight, there is no need to make changes and adjustments according to the characteristics of the indicator, so as to avoid reducing or enlarging certain characteristic values. In order to focus on the comprehensive embodiment of the indicator. Therefore, not only are multiple indicators combined for quality detection, but also the detection efficiency and stability are improved because there is no need to adjust the weights. The accuracy of quality detection is improved.

[0094] Step 107 , based on the image quality detection value sequence, perform image recognition on each remote sensing image in the remote sensing image sequence to generate an image recognition result.

[0095] In some embodiments, the execution subject may perform image recognition on each remote sensing image in the remote sensing image sequence based on the image quality detection value sequence to generate an image recognition result. First, determine the image quality detection value in the image quality detection value sequence that is greater than a preset threshold as a selected quality monitoring value. Then, perform image detection on each remote sensing image corresponding to each selected quality detection value to obtain an image detection result set. Here, the image detection result may include object attributes (e.g., name identification, state identification) and object detection frames detected in the image. Finally, the image detection results in the image detection result set may be fused to obtain an image recognition result. Here, fusion may be a comparison fusion of object attributes and object detection frames corresponding to the same name identification. For example, multiple object detection frames may be averaged for fusion. Multiple object attributes may select the most common attribute as the final attribute in the image recognition result.

[0096] In practice, by identifying remote sensing images whose image quality monitoring values ​​are greater than a preset threshold, images with lower quality can be removed, greatly avoiding the impact of negative samples. Then, through fusion, the features of multiple images can be combined to improve detection accuracy.

[0097] Optionally, the execution subject may also determine the remote sensing images in the remote sensing image sequence corresponding to the image quality detection value that meets the preset detection condition from the image quality detection value sequence as the target remote sensing image, and send each target remote sensing image to the target terminal for display. The preset detection condition may be that the image quality detection value is greater than a preset detection threshold and is the maximum value in the image quality detection value sequence.

[0098] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the remote sensing image recognition method based on multi-index fusion of some embodiments of the present disclosure, the accuracy of remote sensing image recognition can be improved. Specifically, the reason for the reduction in the accuracy of remote sensing image recognition is that the selection of evaluation indicators is not comprehensive enough. Only multiple objective indicators in the surface target evaluation indicators are selected, and it is difficult to comprehensively evaluate the image quality from various aspects. Based on this, the remote sensing image recognition method based on multi-index fusion of some embodiments of the present disclosure first obtains a remote sensing image sequence. Then, each remote sensing image in the above remote sensing image sequence is intercepted to generate a sequence of intercepted remote sensing images. Among them, each intercepted remote sensing image in the above intercepted remote sensing image sequence includes a point target. By intercepting the area where the point target is located, it is convenient to compare the quality of remote sensing images of consecutive frames. Then, the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the above intercepted remote sensing image sequence are determined to obtain a main lobe intensity value sequence and a side lobe intensity value sequence. By introducing the sidelobe intensity value and the mainlobe intensity value, they can be used as the basic data of the point target evaluation index to participate in the index calculation. After that, each of the intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence is transformed to generate a complex autocorrelation feature value sequence. Here, the transformation process can be used to determine the coherence of the image area from the complex level of the pixels in the image, thereby reflecting the evaluation degree of the image in terms of coherence characteristics. Then, based on the above-mentioned mainlobe intensity value sequence and the above-mentioned sidelobe intensity value sequence, each of the intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence is subjected to multi-dimensional index detection to generate an index detection result. Among them, the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category. Here, through multi-dimensional index detection, index detection can be performed from at least the following dimensions: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category, thereby realizing multi-dimensional evaluation. In this way, the comprehensiveness of the index can be improved so that quality detection can be objectively performed from multiple aspects. Finally, based on the above complex autocorrelation eigenvalue sequence and the above index detection results, a comprehensive quality detection is performed on each remote sensing image in the remote sensing image sequence to generate an image quality detection value sequence. Based on the above image quality detection value sequence, image recognition is performed on each remote sensing image in the above remote sensing image sequence to generate an image recognition result. Here, through comprehensive quality detection, multiple dimensional indicators can be used for comprehensive quality evaluation. Therefore, more comprehensive multi-dimensional indicators can be applied to image quality detection to improve the accuracy of remote sensing image quality detection results. In turn, the accuracy of image recognition is improved.

[0099] Further references Figure 2As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a remote sensing image recognition device based on multi-index fusion. These device embodiments are similar to Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0100] like Figure 2 As shown, some embodiments of the remote sensing image recognition device 200 based on multi-index fusion include: an acquisition unit 201, a capture processing unit 202, a determination unit 203, a transformation processing unit 204, a multi-dimensional index detection unit 205 and a comprehensive quality detection unit 206. Among them, the acquisition unit 201 is configured to acquire a remote sensing image sequence; the capture processing unit 202 is configured to perform capture processing on each remote sensing image in the above remote sensing image sequence to generate a captured remote sensing image sequence, wherein each captured remote sensing image in the above captured remote sensing image sequence includes a point target; the determination unit 203 is configured to determine the sidelobe intensity value and the mainlobe intensity value corresponding to the point target included in each captured remote sensing image in the above captured remote sensing image sequence, and obtain a mainlobe intensity value sequence and a sidelobe intensity value sequence; the transformation processing unit 204 is configured to perform transformation processing on each captured remote sensing image in the above captured remote sensing image sequence to generate a captured remote sensing image sequence. Generate a complex autocorrelation eigenvalue sequence; the multidimensional index detection unit 205 is configured to perform multidimensional index detection on each remote sensing image in the above-mentioned intercepted remote sensing image sequence based on the above-mentioned main lobe intensity value sequence and the above-mentioned side lobe intensity value sequence to generate an index detection result, wherein the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category; the comprehensive quality detection unit 206 is configured to perform comprehensive quality detection on each remote sensing image in the remote sensing image sequence based on the above-mentioned complex autocorrelation eigenvalue sequence and the above-mentioned index detection result to generate an image quality detection value sequence. The image recognition unit 207 is configured to perform image recognition on each remote sensing image in the above-mentioned remote sensing image sequence based on the above-mentioned image quality detection value sequence to generate an image recognition result.

[0101] It is understood that the units described in the device 200 are similar to those described in the reference Figure 1 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 200 and the units included therein, and will not be described in detail here.

[0102] Reference below Figure 3 , which shows a structural schematic diagram of an electronic device (eg, a computing device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0103] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory 302 or a program loaded from a storage device 308 into a random access memory 303. Various programs and data required for the operation of the electronic device 300 are also stored in the random access memory 303. The processing device 301, the read-only memory 302, and the random access memory 303 are connected to each other via a bus 304. An input / output interface 305 is also connected to the bus 304.

[0104] Typically, the following devices may be connected to the I / O interface 305: input devices 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 308 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 309. The communication devices 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0105] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the read-only memory 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0106] It should be noted that the computer-readable medium recorded in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0107] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (Hyper Text Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0108] The above-mentioned computer-readable medium may be included in the above-mentioned electronic device; or it may exist independently without being assembled into the electronic device. The above-mentioned computer-readable medium carries one or more programs. When the above-mentioned one or more programs are executed by the electronic device, the electronic device: obtains a remote sensing image sequence; performs interception processing on each remote sensing image in the above-mentioned remote sensing image sequence to generate a sequence of intercepted remote sensing images, wherein each intercepted remote sensing image in the above-mentioned sequence of intercepted remote sensing images includes a point target; determines the sidelobe intensity value and the mainlobe intensity value corresponding to the point target included in each intercepted remote sensing image in the above-mentioned sequence of intercepted remote sensing images, and obtains a mainlobe intensity value sequence and a sidelobe intensity value sequence; performs transformation processing on each intercepted remote sensing image in the above-mentioned sequence of intercepted remote sensing images to generate a sequence of complex autocorrelation eigenvalues. column; based on the main lobe intensity value sequence and the side lobe intensity value sequence, performing multi-dimensional index detection on each of the above-mentioned intercepted remote sensing images in the above-mentioned intercepted remote sensing image sequence to generate an index detection result, wherein the index detection result includes at least one of the following: point target evaluation index category, surface target evaluation index category, information entropy evaluation index category and contrast evaluation index category; based on the above-mentioned complex autocorrelation eigenvalue sequence and the above-mentioned index detection result, performing comprehensive quality detection on each of the remote sensing images in the remote sensing image sequence to generate an image quality detection value sequence; based on the above-mentioned image quality detection value sequence, performing image recognition on each of the remote sensing images in the above-mentioned remote sensing image sequence to generate an image recognition result.

[0109] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0110] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0111] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor including an acquisition unit, an interception processing unit, a determination unit, a transformation processing unit, a multi-dimensional indicator detection unit and a comprehensive quality detection unit, and an image recognition unit. Among them, the names of these units do not constitute a limitation on the units themselves in certain cases, for example, the acquisition unit may also be described as a "unit for acquiring a remote sensing image sequence".

[0112] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0113] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A remote sensing image recognition method based on multi-index fusion, comprising: Acquire remote sensing image sequences; Performing interception processing on each remote sensing image in the remote sensing image sequence to generate an intercepted remote sensing image sequence, wherein each intercepted remote sensing image in the intercepted remote sensing image sequence includes a point target; Determine the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence, and obtain a main lobe intensity value sequence and a side lobe intensity value sequence; Performing transformation processing on each intercepted remote sensing image in the intercepted remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence; Based on the main lobe intensity value sequence and the side lobe intensity value sequence, a multi-dimensional index detection is performed on each intercepted remote sensing image in the intercepted remote sensing image sequence to generate an index detection result, wherein the index detection result includes at least one of the following: a point target evaluation index category, a surface target evaluation index category, an information entropy evaluation index category, and a contrast evaluation index category; Based on the complex autocorrelation eigenvalue sequence and the index detection result, performing comprehensive quality detection on each remote sensing image in the remote sensing image sequence to generate an image quality detection value sequence; Based on the image quality detection value sequence, image recognition is performed on each remote sensing image in the remote sensing image sequence to generate an image recognition result.

2. The method according to claim 1, wherein: The method further comprises: The remote sensing images in the remote sensing image sequence corresponding to the image quality detection values ​​satisfying the preset detection conditions in the image quality detection value sequence are determined as target remote sensing images, and each target remote sensing image is sent to a target terminal for display.

3. The method according to claim 1, wherein: The intercepting and processing each remote sensing image in the remote sensing image sequence to generate an intercepted remote sensing image sequence includes: For each remote sensing image in the remote sensing image sequence, the following steps are performed: Projecting the preset point target coordinates into the remote sensing image to obtain a projected point target coordinate set; Determine the image definition of each position of the projected point target coordinates in the projected point target coordinate set to obtain an image definition set; Selecting the corresponding projected point target coordinates with the highest image clarity from the projected point target coordinate set as the target point position coordinates; Taking the target point position coordinates as the center, image interception is performed according to a preset interception boundary to obtain an intercepted remote sensing image.

4. The method according to claim 3, wherein: The step of determining the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence to obtain a main lobe intensity value sequence and a side lobe intensity value sequence includes: For each intercepted remote sensing image in the intercepted remote sensing image sequence, the following steps are performed: Acquire an impulse response function wave of the coordinate position of the target point in the intercepted remote sensing image, wherein the impulse response function wave includes a main lobe wave and a plurality of side lobe waves; Determine the peak value of the impulse response function wave including the main lobe wave as the main lobe intensity value; The peak value of the largest sidelobe wave among the multiple sidelobe waves included in the impulse response function wave is determined as the sidelobe intensity value.

5. The method according to claim 4, wherein: The transforming process is performed on each of the intercepted remote sensing images in the intercepted remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence, including: For each intercepted remote sensing image in the intercepted remote sensing image sequence, the following steps are performed: A two-dimensional transformation is performed on the complex value of each pixel in the intercepted remote sensing image to generate a complex autocorrelation eigenvalue.

6. The method according to claim 5, wherein: The performing multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result includes: For each intercepted remote sensing image in the intercepted remote sensing image sequence, the following steps are performed: Generate an image peak sidelobe ratio by using the main lobe intensity value and the side lobe intensity value corresponding to the intercepted remote sensing image; Determining the main lobe interval of the intercepted remote sensing image; Determining the image integrated sidelobe ratio of the captured remote sensing image within the mainlobe interval; Determine the ratio of the number of image pixels corresponding to the impulse response function wave corresponding to the intercepted remote sensing image at a preset decibel position to the pixel resolution, and determine the average of each ratio as the spatial resolution index value; The image peak sidelobe ratio, the image integral sidelobe ratio and the spatial resolution index value are determined as index detection results representing the category of point target evaluation indexes.

7. The method according to claim 6, wherein: The method of performing multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result further includes: Gray-scale each of the intercepted remote sensing images in the intercepted remote sensing image sequence to obtain a gray-scale remote sensing image sequence; For each intercepted remote sensing image in the intercepted remote sensing image sequence, the following steps are performed: Determine the pixel intensity average value, pixel intensity variance value and pixel intensity standard deviation of the intercepted remote sensing image by using the length value and width value of the grayscale remote sensing image corresponding to the intercepted remote sensing image; Determine the equivalent visual number of the intercepted remote sensing image by taking the square value of the ratio of the average pixel intensity to the standard deviation of the pixel intensity; Determine the radiometric resolution of the intercepted remote sensing image by using the pixel intensity average value, the pixel intensity standard deviation and the equivalent visual number; The pixel intensity average value, the pixel intensity variance value, the equivalent visual number and the radiation resolution are determined as the indicator detection results that characterize the evaluation category of the surface target.

8. The method according to claim 7, wherein: The method of performing multi-dimensional index detection on each of the intercepted remote sensing images in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result further includes: Determining an information entropy evaluation index value of each intercepted remote sensing image in the intercepted remote sensing image sequence; Based on the grayscale remote sensing image sequence, determining a contrast evaluation index of each intercepted remote sensing image in the intercepted remote sensing image sequence; The information entropy evaluation index value and the contrast evaluation index are determined as index detection results.

9. A remote sensing image recognition device based on multi-index fusion, comprising: An acquisition unit is configured to acquire a remote sensing image sequence; An interception processing unit is configured to perform interception processing on each remote sensing image in the remote sensing image sequence to generate an intercepted remote sensing image sequence, wherein each intercepted remote sensing image in the intercepted remote sensing image sequence includes a point target; A determination unit is configured to determine the side lobe intensity value and the main lobe intensity value corresponding to the point target included in each intercepted remote sensing image in the intercepted remote sensing image sequence, and obtain a main lobe intensity value sequence and a side lobe intensity value sequence; A transformation processing unit is configured to perform transformation processing on each of the intercepted remote sensing images in the intercepted remote sensing image sequence to generate a complex autocorrelation eigenvalue sequence; A multi-dimensional index detection unit is configured to perform multi-dimensional index detection on each remote sensing image in the intercepted remote sensing image sequence based on the main lobe intensity value sequence and the side lobe intensity value sequence to generate an index detection result, wherein the index detection result includes at least one of the following: a point target evaluation index category, a surface target evaluation index category, an information entropy evaluation index category, and a contrast evaluation index category; a comprehensive quality detection unit configured to perform comprehensive quality detection on each remote sensing image in the remote sensing image sequence based on the complex autocorrelation eigenvalue sequence and the index detection result to generate an image quality detection value sequence; The image recognition unit is configured to perform image recognition on each remote sensing image in the remote sensing image sequence based on the image quality detection value sequence to generate an image recognition result.

10. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.

Images