Method, apparatus, computer, and storage medium for enhancing infrared images of the sea surface

Abstract

The present invention discloses a method for enhancing infrared images of the sea surface, including: inputting an infrared image of the sea surface to be enhanced, performing a top-hat transform on the infrared image of the sea surface to be enhanced to obtain a region of interest; adaptively segmenting the region of interest image to obtain a binary labeled image of radiation sources of the infrared image of the sea surface to be enhanced, determining a radiation source region image by using the binary labeled image of radiation sources, and constructing a radiation source suppression image according to the binary labeled image of radiation sources; determining a local illuminance map of the land-air region and a local illuminance map of the sea water region of the infrared image of the sea surface based on the sea horizon; constructing a local guidance map according to the local illuminance map of the land-air region and the local illuminance map of the sea water region to perform background enhancement on the infrared image of the sea surface to be enhanced to obtain a background enhanced image of the infrared image of the sea surface; and superimposing the radiation source region image on the background enhanced image of the infrared image of the sea surface to obtain an enhanced result of the infrared image of the sea surface.

Description

Technical Field

[0001] The present invention belongs to the technical field of image enhancement, and in particular to a method, device, computer, and storage medium for enhancing infrared images of the sea surface. Background Art

[0002] The imaging module of a common border and coastal defense monitoring system is mainly composed of a visible light imaging module and an infrared imaging module. Among them, infrared imaging technology absorbs thermal radiation to generate images, and has the advantages of not relying on visible light conditions and working all-weather. Due to the limitations of the infrared imaging technology principle, the interference of the atmospheric environment, and the attenuation of infrared radiation by the transmission medium, etc., the collected infrared images of the sea surface have problems such as strong spatial correlation, low spatial resolution, low image contrast, and lack of details, which will have an adverse impact on target detection, recognition, analysis, and monitoring. Therefore, improving the contrast and target details of infrared images is an important goal of infrared image enhancement algorithms. In order to enhance the details and contrast of infrared images, a variety of effective infrared image enhancement algorithms have been proposed, which can be divided into the following categories:

[0003] I. Infrared image enhancement algorithms based on histograms. This type of method judges that the image has higher contrast and a larger dynamic range based on the distribution uniformity of the gray levels occupied by pixel points. Using the cumulative distribution function to map the specified input gray level to the output gray level, and then calculating the uniform probability density function, so as to expand the dynamic range of the image gray level and improve the contrast. This type of method only focuses on the global or local image and does not consider the image structure features easily, so image details are lost;

[0004] II. Unsharp masking infrared image enhancement algorithms. The core of this type of algorithm is that the details and contour edges of the image are considered as high-frequency components, and it is hoped to maintain or even limit the low-frequency information while enhancing the high-frequency components. Therefore, a low-pass filter that suppresses the passing of high-frequency components, such as a Gaussian filter, is used to extract the low-frequency component, and the high-frequency component is extracted by subtracting the low-frequency component from the original image. Then, the low-frequency component is constrained and the high-frequency component is appropriately weighted and amplified to achieve enhancing the target details and edges of the image. Common low-pass filters include Gaussian filters, bilateral filters, and guided filters; there are also methods to extract low-frequency and high-frequency components through other transform domains such as the wavelet domain, Curvelet, Contoulet, and shearlet transform domains; the morphological filter with the main idea of geometric decomposition, which decomposes the image in a multi-scale geometric way, can also be regarded as a branch of unsharp masking. This type of enhancement method depends on the selection of spatial filters and gain weight parameters. Traditional low-pass filters will cause overshoot and undersampling when the image is smoothed, resulting in artifacts in the enhanced image; however, the high-frequency components may contain residual noise, and inappropriate gain weights will cause the noise to be amplified and other problems.

[0005] III. Retinex infrared image enhancement algorithm. The model of this type of algorithm believes that the color of an object observed by the human eye is not determined by the visible light reflected into the human eye, but by the inherent reflection coefficient of the object or the scene surface. Here, the visible light reflected into the human eye is considered to be uniform and slowly changing, and it is defined as the illumination component, and the essential information of the target surface is defined as the reflection component. In the calculation method, a low-pass filter is used to estimate the illumination component of the image in the logarithmic domain, and after exponential operation, it returns to the spatial domain to obtain the essential information of the target surface. Although the theory is different from the unsharp masking method, the purpose and calculation method are consistent. The quality of the enhanced image is affected by the filter itself, and it cannot accurately estimate the illumination component and does not have a good edge protection function, resulting in halos at the edges of image details and image distortion.

[0006] In summary, the existing infrared image enhancement methods mainly have the following problems: when the radiation source and the background in the infrared image are enhanced simultaneously, they are prone to mutual influence; gradient inversion is likely to occur at the edges, resulting in artifacts; it is impossible to effectively improve the target details, and it is easy to bring amplified noise, and it is difficult to select the best filter coefficients and weight coefficients. Summary of the Invention

[0007] To solve the above technical problems, an embodiment of the present invention provides a sea surface infrared image enhancement method, which is characterized by including:

[0008] Input the sea surface infrared image to be enhanced, and perform top-hat transformation on the sea surface infrared image to be enhanced to obtain the region of interest;

[0009] Perform adaptive segmentation on the region of interest image to obtain the binary labeling image of the radiation source of the sea surface infrared image to be enhanced, and use the binary labeling image of the radiation source to determine the radiation source region image, and construct a radiation source suppression image according to the binary labeling image of the radiation source;

[0010] Determine the local illumination map of the land-air region and the local illumination map of the sea water region of the sea surface infrared image based on the sea horizon;

[0011] Construct a local guidance map according to the local illumination map of the land-air region and the local illumination map of the sea water region, and perform background enhancement on the sea surface infrared image to be enhanced to obtain the background enhanced map of the sea surface infrared image;

[0012] Overlay the radiation source region image on the background enhanced map of the sea surface infrared to obtain the enhanced result of the sea surface infrared image.

[0013] Further, perform adaptive segmentation on the image of the region of interest to obtain a binary labeling image of radiation sources of the infrared sea surface image to be enhanced, and determine a radiation source region image by using the binary labeling image of radiation sources, including:

[0014] Calculate the histogram of the region of interest;

[0015] Calculate an adaptive segmentation threshold according to the histogram, and segment the image of the region of interest by using the adaptive segmentation threshold to obtain a binary labeling image of radiation sources;

[0016] Multiply the binary labeling image of radiation sources by the image of the region of interest to obtain a radiation source region image.

[0017] Further, calculating the adaptive segmentation threshold according to the histogram includes:

[0018] Perform multivariate non-linear regression fitting on the histogram by using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function;

[0019] Calculate the adaptive segmentation threshold of the image of the region of interest by using the mean and standard deviation of the exponential distribution function.

[0020] Further, determining a local illumination map of the land-air region and a local illumination map of the sea water region of the infrared sea surface image based on the sea horizon includes:

[0021] Determine a land-air labeling map and a sea water labeling map in the infrared sea surface image based on the sea horizon;

[0022] Calculate a land-air region map and a sea water region map in the infrared sea surface image according to the land-air labeling map, the sea water labeling map, and the radiation source suppression image;

[0023] Expand the land-air region map and the sea water region map, and calculate the expanded local illumination map of the land-air region and the local illumination map of the sea water region.

[0024] Further, performing background enhancement on the infrared sea surface image to be enhanced according to the local illumination map of the land-air region and the local illumination map of the sea water region to obtain a background enhanced map of the infrared sea surface image includes:

[0025] Synthesize the local illumination map of the land-air region and the local illumination map of the sea water region to obtain a morphological local illumination map of the infrared sea surface image;

[0026] Calculate a local guidance map for background enhancement of the morphological local illumination map through the radiation source suppression image to obtain the background enhanced map of the infrared sea surface image.

[0027] An infrared image enhancement device for sea surface, comprising:

[0028] An acquisition module, configured to input the infrared image of the sea surface to be enhanced, and perform top-hat transformation on the infrared image of the sea surface to be enhanced to obtain a region of interest;

[0029] A processing module, configured to perform adaptive segmentation on the region-of-interest image to obtain a binary labeled image of radiation sources of the infrared image of the sea surface to be enhanced, determine a radiation source region image by using the binary labeled image of radiation sources, and construct a radiation source suppression image according to the binary labeled image of radiation sources;

[0030] The processing module is configured to determine a local illumination map of the land-air region and a local illumination map of the sea water region of the infrared image of the sea surface based on the sea horizon;

[0031] The processing module is configured to construct a local guidance map according to the local illumination map of the land-air region and the local illumination map of the sea water region, and perform background enhancement on the infrared image of the sea surface to be enhanced to obtain a background enhanced image of the infrared image of the sea surface;

[0032] An execution module, configured to superimpose the radiation source region image on the background enhanced image of the infrared image of the sea surface to obtain an enhancement result of the infrared image of the sea surface.

[0033] Further, the processing module includes:

[0034] A first processing sub-module, configured to calculate a histogram of the region of interest;

[0035] A second processing sub-module, configured to calculate an adaptive segmentation threshold according to the histogram, and perform segmentation on the region-of-interest image by using the adaptive segmentation threshold to obtain a binary labeled image of radiation sources;

[0036] A first execution sub-module, configured to multiply the binary labeled image of radiation sources by the region-of-interest image to obtain a radiation source region image.

[0037] Further, the first processing sub-module includes:

[0038] A first acquisition sub-module, configured to perform multivariate non-linear regression fitting on the histogram by using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function;

[0039] A second acquisition sub-module, configured to calculate an adaptive segmentation threshold of the region-of-interest image by using the mean and standard deviation of the exponential distribution function.

[0040] Further, the processing module includes:

[0041] A third acquisition sub-module, configured to determine a land-air marking map and a sea-water marking map in the sea surface infrared image based on the sea horizon;

[0042] A third processing sub-module, configured to calculate a land-air area map and a sea-water area map in the sea surface infrared image according to the land-air marking map, the sea-water marking map, and the radiation source suppression image;

[0043] A second execution sub-module, configured to expand the land-air area map and the sea-water area map, and calculate a locally illuminated map of the expanded land-air area and a locally illuminated map of the sea-water area.

[0044] Further, the processing module includes:

[0045] A fourth processing sub-module, configured to synthesize the locally illuminated map of the land-air area and the locally illuminated map of the sea-water area to obtain a morphological locally illuminated map of the sea surface infrared image;

[0046] A third execution sub-module, configured to calculate a locally guided map with enhanced background of the morphological locally illuminated map through the radiation source suppression image, and obtain a background enhanced map of the sea surface infrared image.

[0047] A computer device, including a memory and a processor, where computer-readable instructions are stored in the memory, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the above-mentioned sea surface infrared image enhancement method.

[0048] A storage medium storing computer-readable instructions, where when the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the above-mentioned sea surface infrared image enhancement method.

[0049] The beneficial effects of the embodiments of the present invention are as follows: The sea surface infrared image enhancement method based on morphological Retinex local guidance for radiation source reconstruction in the embodiments of the present invention first enhances the radiation source suppression image of the sea surface infrared image, and then reconstructs the radiation source area map in the enhanced image to obtain the final enhanced image, effectively avoiding the problem of inaccurate illuminance estimation caused by radiation source interference, and being able to perform adaptive optimization enhancement on infrared images of different scenes. In addition, morphological filtering is introduced as the center surround function of the Retinex model. By analyzing the structural features of the image after radiation source suppression, land-air and sea-water marking maps are constructed, and a morphological local illuminance estimation method based on marking map constraints is designed to achieve more reliable illuminance estimation. Description of the Drawings

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0051] Figure 1 It is a schematic flowchart of the sea surface infrared image enhancement method provided by the embodiments of the present invention;

[0052] Figure 2 It is a schematic diagram of the sea surface infrared image provided by the embodiments of the present invention;

[0053] Figure 3 It is a schematic diagram of the shape of the sea surface infrared image structuring element provided by the embodiments of the present invention;

[0054] Figure 4 It is a schematic histogram of the region of interest H after top-hat transformation provided by the embodiments of the present invention;

[0055] Figure 5 It is a schematic diagram of the shape of the cross-shaped asymmetric structuring element of the sea surface infrared image provided by the embodiments of the present invention;

[0056] Figure 6 It is a basic structural block diagram of the sea surface infrared image enhancement device provided by the embodiments of the present invention;

[0057] Figure 7 It is a basic structural block diagram of the computer device provided by the embodiments of the present invention. Detailed implementation manners

[0058] In order to enable those skilled in the art to better understand the solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention.

[0059] In some processes described in the specification, claims and above-mentioned drawings of the present invention, there are multiple operations that appear in a specific order. However, it should be clearly understood that these operations can be executed not in the order in which they appear herein or in parallel. The operation numbers such as 101, 102, etc. are only used to distinguish different operations, and the numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations can be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0061] Please refer to Figure 1 , Figure 1 A method for enhancing a sea surface infrared image is provided for an embodiment of the present invention. As Figure 1 shown, the method specifically includes the following steps:

[0062] S1. Input the sea surface infrared image to be enhanced, and perform a top-hat transform on the sea surface infrared image to be enhanced to obtain a region of interest;

[0063] S2. Perform adaptive segmentation on the region of interest image to obtain a binary labeled image of the radiation source of the sea surface infrared image to be enhanced, use the binary labeled image of the radiation source to determine the radiation source region image, and construct a radiation source suppression image according to the binary labeled image of the radiation source;

[0064] S3. Determine the local illumination map of the land-air region and the local illumination map of the sea water region of the sea surface infrared image based on the sea horizon;

[0065] S4. Construct a local guidance map according to the local illumination map of the land-air region and the local illumination map of the sea water region, and perform background enhancement on the sea surface infrared image to be enhanced to obtain a background enhanced map of the sea surface infrared image;

[0066] S5. Superimpose the radiation source region image on the background enhanced map of the sea surface infrared image to obtain the enhanced result of the sea surface infrared image.

[0067] In an embodiment of the present invention, in S1, the sea surface infrared image to be enhanced is obtained, and a top-hat transform is performed on the sea surface infrared image to be enhanced to obtain a region of interest. Specifically:

[0068] Let the sea surface infrared image to be enhanced be I(x, y). As Figure 2 (a) shown, a circular structuring element b1 is selected. As Figure 3 shown, Figure 3 The structuring element shown only represents the general shape of the structuring element. In the Figure 3 example, the radius of the circular structuring element is 10 pixel points. Perform a top-hat transform on the image I(x, y) to be enhanced to obtain a region of interest H:

[0069] I1 = I ° b1

[0070] H(x, y) = I(x, y) - I1(x, y)

[0071] Wherein, ° represents the opening operation operator in morphology, I1 represents the result of the opening operation of the infrared image, H is the top-hat transformation image, and x and y respectively represent the row and column coordinates of the pixel points in the image.

[0072] In one embodiment of the present invention, in step S2, the image of the region of interest is adaptively segmented to obtain the binary labeling image of the radiation source of the infrared image of the sea surface to be enhanced, and the radiation source region image is determined by using the binary labeling image of the radiation source, including:

[0073] Step 1: Calculate the histogram of the region of interest;

[0074] Calculate the histogram hist(H) of the region of interest H after top-hat transformation, as Figure 4 shown.

[0075] Step 2: Calculate the adaptive segmentation threshold according to the histogram, and segment the image of the region of interest by using the adaptive segmentation threshold to obtain the binary labeling image of the radiation source;

[0076] In this embodiment, calculating the adaptive segmentation threshold according to the histogram includes:

[0077] Performing multivariate non-linear regression fitting on the histogram by using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function;

[0078] Calculating the adaptive segmentation threshold of the image of the region of interest by using the mean and standard deviation of the exponential distribution function.

[0079] In this embodiment, using a preset exponential distribution function to perform multivariate non-linear regression fitting on the histogram hist(H), and calculating the corresponding coefficients of the regression equation. The exponential distribution function is as follows:

[0080] p(x) = ae -ax

[0081] In this example, the nlinfit function in MATLAB is used for the corresponding curve fitting, and the parameter a is calculated by fitting. Then, the mean μ and standard deviation σ of the exponential distribution function can be expressed by the parameter a as:

[0082]

[0083]

[0084] Calculating the adaptive segmentation threshold T of the image by using the mean μ and standard deviation σ of the fitted exponential distribution function:

[0085] T = μ + cσ

[0086] Wherein, c is a weighting coefficient, and in actual use, it can be selected to take values in the range of 6 <= c <= 8.

[0087] In this embodiment, the adaptive segmentation threshold is used to segment the image of the region of interest to obtain a binary labeled image of the radiation source, specifically:

[0088] Using the adaptive segmentation threshold T, threshold segmentation is performed on the region of interest H after top-hat transformation to obtain a binary labeled image g(x, y) of the radiation source, as Figure 2 shown in (b).

[0089]

[0090] When the pixel value of a certain point in the region of interest H is greater than the adaptive threshold T, it is determined that this point is a radiation source.

[0091] Step three: Multiply the binary labeled image of the radiation source by the image of the region of interest to obtain a radiation source region image.

[0092] Multiply the binary labeled image of the radiation source by the region of interest image H to obtain a radiation source region image r d :

[0093] r d (x, y) = g(x, y)H(x, y)

[0094] Then, a radiation source suppression image is constructed according to the binary labeled image of the radiation source, specifically: Using the extracted binary labeled image of the radiation source g(x, y) to construct a radiation source suppression image I s :

[0095] S r (x, y) = [1 - g(x, y)]H(x, y) + g(x, y)T

[0096] I s (x, y) = S r (x, y) + I1(x, y)

[0097] Wherein, S r (x, y) represents a clipped image of the radiation source suppression region, and an example image is as Figure 2 shown in (c).

[0098] In an embodiment of the present invention, to overcome the influence of large brightness differences near the sea horizon on contrast estimation, the illumination component L is estimated by combining the sea-sky marked map and morphological closing operation. Specifically, with the sea horizon as a reference, the local illumination maps of the land-air region and the sea-water region of the sea surface infrared image are determined, including:

[0099] Step 1: Determine the land-air marking map and the sea-water marking map in the sea surface infrared image based on the sea horizon line;

[0100] In the embodiments of the present invention, the sea-sky marking map used can be realized by means of existing sea horizon line detection methods. Assume that the sea horizon line equation is as follows:

[0101] Y(x) = kx + b

[0102] Where it is assumed that the size of the infrared image is [N, M], that is, each row of the image has N pixels and each column has M pixels. x represents the column coordinate of the image, that is, x = 1, 2, 3,..., N; round the calculated sea horizon line row coordinate and establish a lookup table array of sea horizon line coordinates:

[0103] Y[x] = [Y(x)] = [kx + b], x = 1, 2,..., N

[0104] Where, [*] represents the rounding operation, and the array Y[x] represents the row coordinate of the corresponding sea horizon line when the column coordinate is x. Store the calculated sea horizon line coordinates in a set of one-dimensional data for convenient lookup.

[0105] Based on the sea horizon line, in each column, the pixel points with row coordinates less than the sea horizon line are classified as the land-air area, and in the land-air marking map M sky set the pixel value at this position to 1, and vice versa, define it as 0; the sea-water marking map and the land-air marking map are complementary to each other. The land-air marking map M sky and the sea-water marking map M sea are specifically defined as follows:

[0106]

[0107] M sea (x, y) = 1 - M sky (x, y)

[0108] Where, for the land-air marking map M sky when M sky = 1, it means that the pixel point is located in the land-air area, and vice versa, it means that the pixel point is located in the sea-water area. Example images of the land-air area and the sea-water area are as shown in Figure 2 (e)-2(f).

[0109] Step 2: Calculate the land-air area map and the sea-water area map in the sea surface infrared image according to the land-air marking map, the sea-water marking map and the radiation source suppression image;

[0110] Through the land-air marking map M sky , the sea-water marking map M sea and the radiation source suppression image Is , calculate the corresponding land-air area map G sky and the sea area map G sea :

[0111] G sky (x, y) = I s (x, y)M sky (x, y)

[0112] G sea (x, y) = I s (x, y)M sea (x, y)

[0113] Step 2: Expand the land-air area map and the sea area map, and calculate the expanded local illumination maps of the land-air area and the sea area.

[0114] To adapt to the morphological operator operation near the sea-sky line, it is necessary to expand the land-air and sea area maps respectively. Taking the sea-sky line as the reference line, select an area with a width of R pixels upward (or downward) as the replication area, and form the expanded area map G through baseline flipping s ' ky and G s ' ea , specifically:

[0115]

[0116]

[0117] Among them, R represents the number of pixels expanded to the area with pixel value 0 (in this embodiment, R = 60 is set), G s ' ky and G s ' ea represent the expanded area maps, as shown in Figure 2 (g)-2(h).

[0118] To calculate the local illumination map constrained by the sea-sky area, it is first necessary to select a morphological structuring element that conforms to the enhancement of the infrared sea surface image. In this embodiment, the infrared sea surface image has the characteristics of horizontal coast and wave distribution, as shown in Figure 5 the cross-shaped asymmetric structuring element b2 shown. Among them, in this embodiment, the major axis of b2 is 60 pixels and the minor axis is 30 pixels,( Figure 5 only represents the shape of the structuring element). Use b2 to calculate the local illumination maps L s ' ky and the sea area G s ' ea respectively, and calculate the local illumination maps L sky and L sea :

[0119] L sky = (G' sky · b2) ∩ M sky

[0120] L sea = (G' sea · b2) ∩ M sea

[0121] Wherein, "·" represents the closing operation in morphology, and "∩" represents the logical intersection operation. L sky (x, y) and L sea (x, y) respectively represent the land-air area illumination map and the sea-water area illumination estimation map.

[0122] Another embodiment of the present invention constructs a local guidance map according to the local illumination map of the land-air area and the local illumination map of the sea-water area, and performs background enhancement on the to-be-enhanced infrared sea surface image to obtain a background enhanced map of the infrared sea surface image, including:

[0123] Step 1: Synthesize the local illumination map of the land-air area and the local illumination map of the sea-water area to obtain a morphological local illumination map of the infrared sea surface image;

[0124] Step 2: Calculate a local guidance map for background enhancement of the morphological local illumination map through the radiation source suppression image to obtain an enhanced map of the radiation source suppression image.

[0125] Synthesize the local illumination maps of the sea and sky regions to obtain a morphological local illumination full map L for radiation source suppression based on the sea-sky marking map:

[0126] L = L sky + L sea

[0127] Wherein, L represents a morphological illumination estimation map realized by radiation source suppression based on the sea-sky marking map, and the result of the illumination estimation example is as Figure 2 (i) shown.

[0128] Calculate the local guidance map R for Retinex background enhancement of the image after radiation source suppression. The calculation formula is as follows:

[0129] r(x, y) = log[I s (x, y)] - log[L(x, y)]

[0130]

[0131] Wherein, log[*] represents the logarithmic operation, exp[*] represents the exponential operation, and R(x, y) represents the calculated local guidance map, as Figure 2(j). Use the local guidance map to enhance the sea surface infrared radiation source suppression image, and obtain the enhanced image I of the sea surface infrared radiation source suppression map c (x, y), such as Figure 2 (k) shown:

[0132] I c (x, y) = I(x, y)R(x, y)

[0133] In the embodiment of the present invention, an adaptive threshold is detected and the radiation source region is extracted through morphological top-hat transformation; then the binary image of the radiation source is marked and the image after suppressing the radiation source is reconstructed; in order to obtain the guidance map for enhancing the background of the sea surface infrared image, the image after suppressing the radiation source is divided into land-air regions and sea water regions by using the sea horizon detection technology, the illuminance of the corresponding regions is estimated respectively by using morphological closing operation, and the two are synthesized to obtain the morphological local illuminance map for suppressing the radiation source based on the sea-sky marking map constraint. Furthermore, the local guidance map for suppressing the radiation source is calculated by using the Retinex method and the morphological local illuminance map, the background information of the sea surface infrared image is enhanced under the guidance of the local guidance map, and the radiation source region is reconstructed in the enhanced image after suppressing the radiation source to obtain the enhanced image of the sea surface infrared image.

[0134] The method for enhancing the sea surface infrared image by reconstructing the radiation source based on morphological Retinex local guidance in the embodiment of the present invention first enhances the background information of the sea surface infrared image, and then reconstructs the radiation source region map in the enhanced image to obtain the final enhanced image, effectively avoiding the problem of inaccurate illuminance estimation caused by radiation source interference, and being able to adaptively optimize and enhance the infrared images of different scenes. In addition, morphological filtering is introduced as the center surround function of the Retinex model. By analyzing the structural characteristics of the image after suppressing the radiation source, land-air and sea water marking maps are constructed, and a morphological local illuminance estimation method based on the marking map constraint is designed to achieve more reliable illuminance estimation.

[0135] Such as Figure 6As shown in the figure, to solve the above problems, an embodiment of the present invention further provides a sea surface infrared image enhancement device, including: an acquisition module 2100, a processing module 2200, and an execution module 2300. Among them, the acquisition module 2100 is used to input the sea surface infrared image to be enhanced, and perform a top-hat transform on the sea surface infrared image to be enhanced to obtain a region of interest; the processing module 2200 is used to adaptively segment the region of interest image to obtain a binary labeled image of the radiation source of the sea surface infrared image to be enhanced, and use the binary labeled image of the radiation source to determine a radiation source region image, and construct a radiation source suppression image according to the binary labeled image of the radiation source; the processing module 2200 is used to determine a local illumination map of the land-air region and a local illumination map of the sea water region of the sea surface infrared image based on the sea horizon; the processing module 2200 is used to construct a local guidance map according to the local illumination map of the land-air region and the local illumination map of the sea water region to perform background enhancement on the sea surface infrared image to be enhanced to obtain a background enhanced map of the sea surface infrared image; the execution module 2300 is used to superimpose the radiation source region image on the enhanced map of the radiation source suppression image to obtain an enhanced result of the sea surface infrared image.

[0136] The sea surface infrared image enhancement method based on morphological Retinex local guidance radiation source reconstruction in the embodiment of the present invention first enhances the radiation source suppression image of the sea surface infrared image, and then reconstructs the radiation source region map in the enhanced image to obtain the final enhanced image, effectively avoiding the problem of inaccurate illumination estimation caused by radiation source interference, and being able to perform adaptive optimization enhancement on infrared images of different scenes. In addition, morphological filtering is introduced as the center surround function of the Retinex model. By analyzing the structural characteristics of the image after radiation source suppression, a land-air and sea water labeled map is constructed, and a morphological local illumination estimation method based on the labeled map constraint is designed to achieve more reliable illumination estimation.

[0137] In some embodiments, the processing module includes: a first processing sub-module, configured to calculate a histogram of the region of interest; a second processing sub-module, configured to calculate an adaptive segmentation threshold according to the histogram, and segment the region of interest image using the adaptive segmentation threshold to obtain a binary labeled image of the radiation source; a first execution sub-module, configured to multiply the binary labeled image of the radiation source by the region of interest image to obtain a radiation source region image.

[0138] In some embodiments, the first processing sub-module includes: a first acquisition sub-module, configured to perform multivariate non-linear regression fitting on the histogram using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function; a second acquisition sub-module, configured to calculate the adaptive segmentation threshold of the region of interest image using the mean and standard deviation of the exponential distribution function.

[0139] In some embodiments, the processing module includes: a third acquisition sub-module, configured to determine a land-air marker map and a sea-water marker map in the sea surface infrared image based on the sea horizon; a third processing sub-module, configured to calculate a land-air area map and a sea-water area map in the sea surface infrared image according to the land-air marker map, the sea-water marker map, and the radiation source suppression image; a second execution sub-module, configured to expand the land-air area map and the sea-water area map, and calculate a local illuminance map of the expanded land-air area and a local illuminance map of the sea-water area.

[0140] In some embodiments, the processing module includes: a fourth processing sub-module, configured to synthesize the local illuminance map of the land-air area and the local illuminance map of the sea-water area to obtain a morphological local illuminance map of the sea surface infrared image; a third execution sub-module, configured to calculate a local guidance map with enhanced background of the morphological local illuminance map through the radiation source suppression image, to obtain a background enhanced map of the sea surface infrared image.

[0141] To solve the above technical problems, an embodiment of the present invention further provides a computer device. Specifically, please refer to Figure 7 , Figure 7 which is a basic structural block diagram of the computer device in this embodiment.

[0142] As Figure 7 shown, it is a schematic internal structure diagram of the computer device. As Figure 7 shown, the computer device includes a processor, a non-volatile storage medium, a memory, and a network interface connected through a system bus. Among them, the non-volatile storage medium of the computer device stores an operating system, a database, and computer-readable instructions. The database can store a control information sequence. When the computer-readable instructions are executed by the processor, the processor can implement an image processing method. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The memory of the computer device can store computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor can execute an image processing method. The network interface of the computer device is used to connect and communicate with a terminal. Those skilled in the art can understand that Figure 7 the structure shown in

[0143] In this embodiment, the processor is used to execute Figure 6The specific contents of the acquisition module 2100, the processing module 2200, and the execution module 2300 are obtained. The memory stores the program codes and various types of data required to execute the above modules. The network interface is used for data transmission between the user terminal and the server. In this embodiment, the memory stores the program codes and data required to execute all sub-modules in the image processing method, and the server can call the program codes and data of the server to execute the functions of all sub-modules.

[0144] The computer device provided by the embodiment of the present invention, the method for enhancing the infrared image of the sea surface based on the radiation source reconstruction guided by morphological Retinex in the embodiment of the present invention, first enhances the radiation source suppression image of the infrared image of the sea surface, and then reconstructs the radiation source area map in the enhanced image to obtain the final enhanced image, effectively avoiding the problem of inaccurate illuminance estimation caused by radiation source interference, and being able to adaptively optimize and enhance the infrared images of different scenes. In addition, morphological filtering is introduced as the center surround function of the Retinex model. By analyzing the structural characteristics of the image after radiation source suppression, land-air and seawater marker maps are constructed, and a morphological local illuminance estimation method based on marker map constraints is designed to achieve more reliable illuminance estimation.

[0145] The present invention also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to execute the steps of the image processing method described in any of the above embodiments.

[0146] 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 computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), etc., or a random access memory (RAM), etc.

[0147] It should be understood that although the steps in the flowchart of the accompanying drawings are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. Their execution order is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0148] The above are only some embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. An infrared image enhancement method for the sea surface, characterized in that Including: Input the infrared image of the sea surface to be enhanced, and perform top-hat transformation on the infrared image of the sea surface to be enhanced to obtain the region of interest; Perform adaptive segmentation on the region of interest image to obtain the binary labeling image of the radiation source of the infrared image of the sea surface to be enhanced, use the binary labeling image of the radiation source to determine the radiation source region image, and construct a radiation source suppression image according to the binary labeling image of the radiation source; Determine the local illumination map of the land-air region and the local illumination map of the sea water region of the infrared image of the sea surface based on the sea horizon; Construct a local guidance map according to the local illumination map of the land-air region and the local illumination map of the sea water region to perform background enhancement on the infrared image of the sea surface to be enhanced to obtain the background enhanced map of the infrared image of the sea surface; Overlay the radiation source region image on the background enhanced map of the infrared image of the sea surface to obtain the enhanced result of the infrared image of the sea surface.

2. The method according to claim 1, wherein Performing adaptive segmentation on the region of interest image to obtain the binary labeling image of the radiation source of the infrared image of the sea surface to be enhanced, and using the binary labeling image of the radiation source to determine the radiation source region image, including: Calculate the histogram of the region of interest; Calculate the adaptive segmentation threshold according to the histogram, and use the adaptive segmentation threshold to segment the region of interest image to obtain the binary labeling image of the radiation source; Multiply the binary labeling image of the radiation source by the region of interest image to obtain the radiation source region image.

3. The method according to claim 2, characterized in that, Calculating the adaptive segmentation threshold according to the histogram, including: Perform multivariate non-linear regression fitting on the histogram using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function; Calculate the adaptive segmentation threshold of the region of interest image using the mean and standard deviation of the exponential distribution function.

4. The method according to claim 1, wherein Determining the local illumination map of the land-air region and the local illumination map of the sea water region of the infrared image of the sea surface based on the sea horizon, including: Determine the land-air labeling map and the sea water labeling map in the infrared image of the sea surface based on the sea horizon; Calculate the land-air region map and the sea water region map in the infrared image of the sea surface according to the land-air labeling map, the sea water labeling map and the radiation source suppression image; Expand the land-air region map and the sea water region map, and calculate the expanded local illumination map of the land-air region and the local illumination map of the sea water region.

5. The method according to claim 1, wherein Performing suppression enhancement on the infrared image of the sea surface to be enhanced according to the local illumination map of the land-air region and the local illumination map of the sea water region to obtain the background enhanced map of the infrared image of the sea surface, including: Synthesize the local illumination map of the land-air region and the local illumination map of the sea water region to obtain the morphological local illumination map of the infrared image of the sea surface; Calculate the local guidance map for background enhancement of the morphological local illumination map through the radiation source suppression image to obtain the background enhanced map of the infrared image of the sea surface.

6. An infrared image enhancement device for sea surface, characterized in that, Including: An acquisition module, configured to input the infrared image of the sea surface to be enhanced, and perform top-hat transformation on the infrared image of the sea surface to be enhanced to obtain the region of interest; A processing module, configured to adaptively segment the image of the region of interest to obtain a binary labeled image of radiation sources of the sea surface infrared image to be enhanced, determine a radiation source region image by using the binary labeled image of radiation sources, and construct a radiation source suppression image according to the binary labeled image of radiation sources; The processing module is configured to determine a local illuminance map of the land-air region and a local illuminance map of the sea water region of the sea surface infrared image based on the sea horizon to construct a local guidance map; The processing module is configured to perform background enhancement on the sea surface infrared image to be enhanced by using the local guidance map constructed according to the local illuminance map of the land-air region and the local illuminance map of the sea water region to obtain a background enhanced map of the sea surface infrared image; An execution module, configured to superimpose the radiation source region image on the background enhanced map of the sea surface infrared image to obtain an enhancement result of the sea surface infrared image.

7. The device according to claim 6, characterized in that, The processing module includes: A first processing sub-module, configured to calculate a histogram of the region of interest; A second processing sub-module, configured to calculate an adaptive segmentation threshold according to the histogram, and segment the image of the region of interest by using the adaptive segmentation threshold to obtain a binary labeled image of radiation sources; A first execution sub-module, configured to multiply the binary labeled image of radiation sources by the image of the region of interest to obtain a radiation source region image.

8. The device according to claim 7, wherein The first processing sub-module includes: A first acquisition sub-module, configured to perform multivariate non-linear regression fitting on the histogram by using a preset exponential distribution function to obtain the mean and standard deviation of the exponential distribution function; A second acquisition sub-module, configured to calculate an adaptive segmentation threshold of the image of the region of interest by using the mean and standard deviation of the exponential distribution function.

9. A computer device, comprising a memory and a processor, wherein computer-readable instructions are stored in the memory, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the sea surface infrared image enhancement method according to any one of claims 1 to 5.

10. A storage medium storing computer-readable instructions, wherein when the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the sea surface infrared image enhancement method according to any one of claims 1 to 5.

Images