High-resolution three-spectral band rate thermal infrared aerial plume enhancement method based on brightness temperature characteristics

Abstract

The application discloses a high-resolution three-spectral band rate thermal infrared aerial wake enhancement method based on brightness temperature characteristics, and comprises the following steps: (1) cutting a satellite thermal infrared image containing an aerial wake into a uniform size subgraph; (2) for each subgraph, generating a wake enhancement mask according to the brightness temperature characteristics of the wake; (3) performing percentage cut-off linear stretching processing on each subgraph; (4) for the stretched image generated in the step (3), enhancing the image saturation in the wake enhancement mask region according to the wake enhancement mask generated in the step (2), and obtaining a wake enhancement image. The method of the application significantly improves the detection precision and image quality of the wake in the high-resolution three-spectral band thermal infrared image, and has wide application value and development prospect.

Description

Technical Field

[0001] This invention relates to the field of satellite detection and image processing technology, and in particular to a high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics. Background Technology

[0002] Aircraft flying at high altitudes typically create condensation contrails due to the influence of cold air. In recent years, with the rapid development of aerospace technology, aircraft contrails have significantly impacted global climate change. Furthermore, although contrails do not affect aircraft performance, they can reveal the location, number, and movement of aircraft, making reverse tracking of aircraft direction and movement equally important. Therefore, accurate contrail observation is crucial in multiple fields, including environmental monitoring, aviation safety, national security, and air traffic management.

[0003] At present, the methods for detecting and enhancing aircraft contrails mainly rely on remote sensing technologies, including optical and thermal infrared imaging. Thermal infrared (TIR) ​​remote sensing, a long-standing tool in environmental monitoring, has been continuously developed. Its stability and day / night observation capabilities provide a solid foundation for all-weather, uninterrupted contrail observation. Therefore, TIR technology has become one of the primary means of observing aircraft contrails.

[0004] However, a major limitation of current thermal infrared remote sensing technology is its relatively low spatial resolution. This means that while the presence of aircraft contrails can be detected, the identification and analysis of details such as their precise shape, size, or complex interactions with the atmosphere are limited. The information loss caused by low resolution makes comprehensive and detailed detection of contrails a challenge.

[0005] Launched on November 5, 2021, SDGSAT-1 represents a significant milestone as the first scientific satellite designed to support the United Nations 2030 Sustainable Development Goals, capable of global imaging every two weeks. The satellite is uniquely equipped with a Thermal Infrared Spectrometer (TIS) offering high resolution and wide swath capabilities. The TIS provides a fine resolution of 30 meters across three spectral bands: 8–10.5 μm, 10.3–11.3 μm, and 11.5–12.5 μm, redefining the standard for high-resolution thermal infrared satellite payloads. These innovative features highlight its unique contribution to environmental monitoring and detailed spatial analysis, demonstrating the broad application potential of thermal infrared remote sensing technology in future environmental science research.

[0006] Advances in TIS technology are crucial for understanding and monitoring atmospheric phenomena and the impacts of human activities. First, three-band false-color imagery helps highlight contrails. Contrails and natural cirrus clouds exhibit significant differences in shape and physical properties. Contrails initially appear as elongated straight lines, while natural cirrus clouds display complex, feathery structures. Over time, the shape of contrails changes due to environmental factors such as humidity and wind direction, while the composition of natural cirrus clouds remains more stable. Furthermore, unlike natural cirrus clouds, contrails are composed of smaller, randomly oriented ice crystals, forming high-altitude ice clouds.

[0007] Using the Thermal Infrared Spectrometer (TIS) on the SDGSAT-1, and leveraging the differences in thermal radiation across three spectral bands (Band 1: 8–10.5 μm, Band 2: 10.3–11.3 μm, Band 3: 11.5–12.5 μm), precise three-band false-color images can be generated. These bands provide crucial details about surface or cloud top temperature, cloud type, and temperature differences with the ground. Band 1 (B1) provides temperature insights through its strong radiative penetrating power. Band 2 (B2) effectively distinguishes between water and ice clouds, revealing the cloud's physical state by reflecting the absorption characteristics of water molecules. Band 3 (B3) focuses on the temperature difference between the ground and cloud top, thus enhancing the distinction between contrails and cirrus clouds in the false-color image.

[0008] Secondly, the 30-meter high resolution facilitates detailed contrail detection. The Thermal Infrared Spectrometer (TIS) on SDGSAT-1, with its three spectral bands, provides a fine-grained spatial resolution of 30 meters, setting a new standard for high-resolution thermal infrared satellite payloads. This advanced imaging technology offers two key enhancements in the accuracy and detail resolution of contrail detection. First, capturing image information of the aircraft's nose plays a crucial role in auxiliary verification to determine whether an observed cloud is a contrail. The 30-meter high-resolution thermal infrared image reveals minute details between the aircraft and its contrail, which are often difficult to discern in low-resolution images. By identifying the thermal radiation characteristics of the aircraft's nose, we can not only confirm that the observed contrail was produced by a specific aircraft but also significantly improve the accuracy of distinguishing contrails from natural cirrus clouds. Second, high-resolution imaging technology enhances the ability to detect faint contrails, which are often overlooked in low-resolution images due to insufficient detail and thermal intensity. This capability not only facilitates the comprehensive detection and recording of all contrails but also provides deeper insights into their formation and evolution, contributing to a more thorough understanding of contrail impact.

[0009] Meanwhile, the shape of the contrail is more subtle than that of other natural clouds. Therefore, before formal detection, making full use of the characteristics of three-band thermal infrared remote sensing to preprocess the image to enhance the contrail, thereby improving the detection efficiency, is one of the important steps in the detection process.

[0010] Existing methods for enhancing wakes in thermal infrared remote sensing include generating spectral band difference false-color images and Hough line detection. However, these methods primarily analyze and process long, prominent wakes in low-resolution images. They are not suitable for high-resolution images and neglect the enhancement of subtle wakes, thus affecting the comprehensive detection and analysis of wakes.

[0011] Therefore, developing and implementing an enhancement method for high-resolution three-band thermal infrared remote sensing aircraft contrails is an urgent problem to be solved. Summary of the Invention

[0012] This invention provides a high-resolution three-band thermal infrared contrail enhancement method based on brightness temperature characteristics, which can enhance both long and short contrails, improving image visualization quality and analysis efficiency.

[0013] The technical solution of the present invention is as follows: A high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics includes the following steps: (1) Cropping satellite thermal infrared images containing flight contrails into sub-images of uniform size; (2) For each sub-image, generate a trail enhancement mask based on the brightness temperature characteristics of the trail; (3) Perform percentage-based truncation and linear stretching on each subgraph; (4) For the stretched image generated in step (3), the image saturation in the area of ​​the trail enhancement mask generated in step (2) is enhanced to obtain the trail enhancement image.

[0014] Step (2) includes: (2-1) For each sub-image, calculate the brightness temperature value of each pixel in the sub-image in multiple spectral bands and the brightness temperature ratio between multiple spectral bands; (2-2) For each sub-image, the brightness temperature value and brightness temperature ratio are sorted by size. Based on the brightness temperature size characteristics and brightness temperature ratio characteristics of the flight trail, the brightness temperature value threshold and brightness temperature ratio threshold based on percentiles are selected. (2-3) Define the mask according to the selected brightness temperature threshold and brightness temperature ratio threshold respectively.

[0015] Taking the SDGSAT-1 satellite as an example, the remote sensing images used have three spectral bands: the center wavelength of the B1 band is 9.35 μm, the center wavelength of the B2 band is 10.73 μm, and the center wavelength of the B3 band is 11.72 μm.

[0016] In step (2-1), the formula for calculating the brightness temperature value is: L i = GAINi •DN i +BLAS i , i =1,2,3 Where: L i ( i =1,2,3) is B i Radiance of the spectral band; h It is Planck's constant. c It's the speed of light. k It is Boltzmann's constant. λ It is the center wavelength. k 1 and k 2 is a constant; T i ( i =1,2,3) B i Brightness temperature value of the spectral band; GAIN i ( i =1,2,3) is B i The calibration gain coefficient of the spectral band, DN i ( i =1,2,3) is B i The effective recorded value of a spectral pixel. BIAS i ( i =1,2,3) is B i The calibration offset of the spectral band.

[0017] Furthermore, taking SDGSAT-1 satellite imagery as an example, GAIN i , DN i , BIAS i As shown in Table 1: Table 1 .

[0018] In step (2-1), the formula for calculating the brightness-temperature ratio is: in: T 1 represents the brightness temperature of the B1 spectral band. T 2 represents the brightness temperature of the B2 spectral band. T3 indicates the brightness temperature of the B3 spectral band; R 1. R 2. R 3 represents the ratio of the brightness temperature of the B1 band to the brightness temperature of the B3 band, the ratio of the brightness temperature of the B2 band to the brightness temperature of the B3 band, and the ratio of the brightness temperature of the B2 band to the brightness temperature of the B1 band, respectively.

[0019] Step (2-2) includes: (2-2i) Sort the brightness temperature values ​​by magnitude: in: For B i The order of brightness temperature values ​​of the spectral bands T i (n) represents B i The brightness temperature values ​​of the spectral bands are sorted by size. n The element at position, n This represents the total number of pixels in the sub-image. (2-2ii) Set the brightness temperature threshold value according to the brightness temperature characteristics of the flight contrail, using the following formula: in: t It is the B that is being filtered. i The percentage position of the brightness temperature values ​​in the spectral bands; TH Ti It is the B that is being filtered. i Brightness temperature threshold for the spectral band; n This represents the total number of pixels in the sub-image. Indicates to t Round down; (2-2iii) Sort the brightness-temperature ratios by magnitude: in: Brightness-to-temperature ratio R j The sorting; R j ( n )for The Middle n The element at the specified position; n This represents the total number of pixels in the sub-image. (2-2iv) Set the brightness temperature ratio threshold based on the brightness temperature characteristics of the flight contrail: in: TH Rj It is a filter. The brightness temperature ratio threshold; r j It is a filter. The percentage position in the text.

[0020] Steps (2-3) include: (2-3i) Two types of masks are defined based on the brightness temperature threshold and the brightness temperature ratio threshold: The mask for the wake enhancement region is calculated using the following formula. Mask : .

[0021] Step (3) includes: (3-1) Calculate the 0% and 95% percentiles of each pixel in the sub-image; (3-2) Using the pixel values ​​at the 0th and 95th percentiles as the lower and upper limits of stretching, respectively, the pixel values ​​of the sub-image are linearly stretched to generate the stretched image.

[0022] Step (4) includes: (4-1) Convert the stretched image from an RGB image to an HSV image; (4-2) Based on the trail enhancement mask generated in step (2), the saturation in the HSV image is enhanced using the following formula: in: S original This represents the original saturation value of any pixel within the trail enhancement mask. s >1 is a predetermined coefficient for saturation enhancement. S Enhanced This is the enhanced saturation value; (4-3) Merge the enhanced saturation and brightness back into the HSV image, and then convert it back into an RGB image.

[0023] This invention first crops the original satellite image to achieve targeted and precise processing of the contrails. Second, by deeply analyzing the physical and infrared imaging characteristics of the contrails, a mask for contrail enhancement is generated, enabling the selection of the enhancement area. Next, the cropped image is truncated and stretched by a percentage, effectively reducing or eliminating extreme grayscale values ​​typically caused by noise, cloud cover, or image edge effects, which can interfere with the accuracy and efficiency of contrail enhancement. Finally, the saturation of the image is extracted and enhanced using RGB to HSV color space conversion, achieving effective enhancement of aviation contrails while preserving image detail features.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows: First, by analyzing the physical principles and applying the brightness temperature characteristics of contrails, and then setting the brightness temperature magnitude of each spectral band and the percentile threshold of the brightness temperature ratio between spectral bands, it is possible to accurately identify and enhance aviation contrails in high-resolution thermal infrared images, significantly improving the visibility and accuracy of contrails analysis. Second, in terms of image processing, the percentage-truncated linear stretching and saturation enhancement techniques employed in this invention significantly improve the visual effect of the image, making the contrails more prominent and facilitating further detection and analysis.

[0025] The method of this invention significantly improves the detection accuracy and image quality of wakes in high-resolution three-band thermal infrared images, and has broad application value and development prospects. Attached Figure Description

[0026] Figure 1 This is a flowchart of the high-resolution three-band thermal infrared aircraft contrail enhancement method based on brightness temperature characteristics according to the present invention. Figure 2 A diagram illustrating how to crop a satellite image into a smaller version; Figure 3 The flowchart for step 3 is as follows: generating a wake enhancement mask based on the wake brightness temperature characteristics. Figure 4 The graph shows the reflectivity of ice and water as a function of wavelength. Figure 5 A percentage frequency distribution of brightness temperature values ​​across the three spectral segments of the wake; Figure 6 A percentage position frequency statistical graph showing the magnitude of the brightness temperature ratio among the three spectral segments of the wake; Figure 7 The illustration shows an example of wake enhancement using the method of the present invention, where (a) is a short wake and (b) is a long wake. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the embodiments described below are intended to facilitate the understanding of the present invention and do not limit it in any way.

[0028] This invention aims to address the problems in existing technologies by proposing a high-resolution three-band thermal infrared aircraft contrails enhancement method based on brightness-temperature characteristics. Through a series of innovative technical steps, this invention effectively enhances aircraft contrails in thermal infrared remote sensing images, improving image visualization quality and analysis efficiency.

[0029] The specific technical solution includes successively cropping the original image to generate more detailed sub-images, setting brightness temperature ratio thresholds between different spectral bands and brightness temperature magnitude thresholds for each spectral band based on the thermal infrared brightness temperature characteristics of the wake to generate a wake enhancement mask, performing percentage truncation and linear stretching on the cropped image to obtain an 8-bit image, and finally improving the saturation of the wake mask region through saturation enhancement to obtain an enhanced image.

[0030] like Figure 1 As shown, the present invention provides a high-resolution three-band thermal infrared aircraft contrail enhancement method based on brightness-temperature characteristics, comprising the following steps: S1, gradually crop the original image along the horizontal and vertical directions to generate a small image of uniform size.

[0031] Step S1 involves cropping the original image to generate a cropped image. Specifically, as follows... Figure 2 As shown, the starting point is first determined from the top left corner of the original large satellite image. Then, the image is gradually cropped by moving along the horizontal and vertical directions at a predetermined small image size. During this process, the cropping operation follows a left-to-right, then top-to-bottom order to ensure complete coverage of every part of the original large image, and that the size of the small images obtained from each cropping remains consistent. Finally, all small images are uniformly encoded and stored for subsequent processing and analysis. In this invention, the size of the small image is 256. 256.

[0032] S2, For the cropped image obtained in step S1, generate a trail enhancement mask based on the trail brightness temperature characteristics, such as... Figure 3 As shown, it includes the following steps: S2.1 For each cropped small image, calculate the brightness temperature value of each pixel in the image in the three spectral bands and the brightness temperature ratio between the three spectral bands; First, the brightness temperature is calculated using the following formula: L i = GAIN i •DN i +BLAS i , i =1,2,3(1) (2) (3) Equation (3) is the formula for calculating the brightness temperature T. This invention uses the SDGSAT-1 satellite as an example, employing remote sensing images with three spectral bands: the center wavelength of band B1 is 9.35 μm, the center wavelength of band B2 is 10.73 μm, and the center wavelength of band B3 is 11.72 μm. (Constant) k 1 andk 2. Based on Planck's radiation law and Wien's displacement law, a formula is derived to convert the received brightness into brightness temperature, as shown in equation (2). Wherein, h It is Planck's constant. c It's the speed of light. k It is Boltzmann's constant. λ It is the center wavelength. L i ( i =1,2,3) represent the radiance in the 9.35μm, 10.73μm, and 11.72μm bands, respectively, as shown in equation (1). Wherein, GAIN The calibration gain coefficient representing a specific spectral band. DN This represents the effective recorded value of the pixel in this spectral band. BIAS This is the calibration offset for this spectral band, and the calibration coefficients are shown in Table 1.

[0033] Table 1 SDGSAT-1 TIS Radiation Calibration Factors Secondly, the brightness temperature ratio of each pixel across the three spectral bands is calculated. This step enhances the contrast between the trail and the background by utilizing the different thermal properties of the trail at different spectral wavelengths. The calculation formula is: (4) in, T 1 represents the brightness temperature of the B1 spectral band. T 2 represents the brightness temperature of the B2 spectral band. T 3 indicates the brightness temperature of the B3 spectral band; R 1. R 2. R 3 refers to the ratio of the brightness temperature of the B1 band to the brightness temperature of the B3 band, the ratio of the brightness temperature of the B2 band to the brightness temperature of the B3 band, and the ratio of the brightness temperature of the B2 band to the brightness temperature of the B1 band, respectively.

[0034] S2.2 In each cropped small image, the brightness temperature value and brightness temperature ratio are sorted by size, and a threshold based on percentiles is selected according to the brightness temperature size characteristics and brightness temperature ratio characteristics of the trail. The physical principle upon which this step is based is: Contrails typically appear at altitudes above 8000 meters and are characterized by low temperatures, resulting in a temperature difference of 4 to 0.6 K between the contrail and the cloud background. This creates a negative contrast with ground objects in infrared remote sensing images. Compared to natural clouds, contrails are primarily composed of small, randomly oriented ice crystals. Figure 4(The image shows the reflectivity characteristics of ice and seawater within the SDGSAT-1 TIS wavelength range.) Higher reflectivity indicates weaker absorption, and therefore a lower brightness temperature. Analysis revealed that the brightness temperature differences between the B2 and B1 bands, and between the B2 and B3 bands, are crucial for distinguishing contrails from background and natural clouds. Regarding the brightness temperature difference between the B2 and B1 bands, ice has a lower brightness temperature than seawater; within the B2–B3 band range, the brightness temperature of ice decreases with increasing wavelength, due to the variation in absorption characteristics of ice crystals with wavelength. For smaller particles, due to stronger reflection and weaker absorption at longer wavelengths, the brightness temperature decreases more rapidly with wavelength, exhibiting a steeper slope.

[0035] In this step, it is necessary to calculate the brightness temperature percentile (Pct_BT) and brightness temperature ratio percentile (Pct_BTR) of the trail in each 256*256 small image.

[0036] Percentiles are a statistical measure representing the percentage of observations below a certain value. Using percentiles helps standardize subsequent thresholds, making the method more adaptable to different background conditions and enhancing its versatility. The Pct_BT and Pct_BTR of the trails were analyzed and statistically tested. Considering that the trail in region B1 has low contrast with the background and is easily confused, the Pct_BT of regions B2 and B3 was statistically analyzed because the trails in these two regions are easier to distinguish. (See Figure 5.) Figure 6 As shown, the brightness temperature values ​​of the wakes in thermal infrared images are generally lower than 90% of Pct_BT. The ratios of B1 to B3 and B2 to B3 are relatively high, usually exceeding the 40th percentile_BTR, while the ratio of B2 to B1 is often lower, usually below the 50th percentile_BTR.

[0037] Regarding the calculation process, firstly, the brightness temperatures obtained in step S2.1 are sorted from smallest to largest, as shown in the following formula: (5) in, T i (n) represents the nth i The brightness temperature values ​​of the spectral bands are sorted by size. n The element at position, n It is the total number of pixels in each 256*256 image. It is sorted by the brightness temperature values ​​calculated earlier.

[0038] Next, a threshold is set based on the brightness temperature characteristics of the wake. The calculation formula is as follows: (6) in, t It is to screen B i Percentage position of brightness temperature sorting in spectral bands; TH Ti It is to screen B i Brightness temperature threshold of the spectral band n It is the total number of pixels in each 256*256 image. Indicates to t Round down to the nearest integer.

[0039] A trail enhancement mask based on percentiles sorted by brightness temperature is defined. By comparing the brightness temperature of each spectral band with its 90th percentile, it is determined which regions have temperatures above a threshold.

[0040] Secondly, the brightness temperature ratios between spectral bands obtained in step S2.1 are sorted from smallest to largest, as shown in the following formula: (7) in, R j ( n ) represents the first j In the ranking of brightness temperature ratios between spectral bands, the first n The element at position, n It is the total number of pixels in each 256*256 image. This sorts the brightness temperature ratios between the previously calculated spectral bands.

[0041] Next, a threshold is set based on the brightness temperature characteristics of the wake. The calculation formula is as follows: (8) in, r j It is the first screening j Brightness temperature ratio between spectral segments R j The percentage position in the sorting; TH Rj It is the first screening j Threshold for sorting brightness temperature ratios among spectral bands.

[0042] S2.3, Define two different masks based on the selected threshold to identify the trail region in the image; First type of mask ( Mask 1) Aimed at identifying wake regions with lower brightness temperatures. The calculation formula is as follows: (9) The second type of mask ( Mask2) Identify wheezes in images by setting a specific brightness-to-temperature ratio threshold. The calculation formula is as follows: Finally, the mask for the wake enhancement region is defined as the product of Mask1 and Mask2, calculated as follows: (10) S3, For the cropped small image obtained in step S1, the cropped image is processed by percentage truncation and linear stretching to generate an 8-bit depth image; In this step, we apply a percentage-truncation linear stretching method to the small images cropped during preprocessing to obtain 8-bit depth images. This method aims to improve image quality and make the wake easier to distinguish. The key is to adjust the image contrast by linearly stretching the grayscale range, effectively reducing or eliminating extreme grayscale values ​​that are usually caused by noise, cloud cover, or image edge effects, which can interfere with the accuracy and efficiency of wake enhancement. Considering the low temperature of the wake, resulting in a generally low DN value distribution, we adjust the DN value range to a stretching range between 0-95%.

[0043] To implement this method, the 0% and 95th percentiles of the pixels in the image are first calculated as the lower and upper limits of the stretching. This is done to remove extreme grayscale values, which typically affect the contrast and detail of the image.

[0044] Next, a truncation operation is performed on the image portion to remove outliers by limiting pixel values ​​to upper and lower limits. Following this, a linear stretching process is performed, linearly mapping the truncated pixel values ​​to a specified output range. This mapping process adjusts the image contrast, allowing DN values ​​that were originally distributed within a narrow range to be displayed over a wider output range, thereby enhancing image recognizability and detail.

[0045] Finally, by assigning the processed pixel values ​​back to the corresponding channels, the percentage-truncated linear stretching of the image is completed. This method not only improves image quality but also makes details such as wheezes at low temperatures more apparent by effectively adjusting image contrast, thus helping to improve the accuracy and efficiency of wheez enhancement.

[0046] S4. The 8-bit depth image obtained in step S3 and the mask obtained in step S2 are used to enhance the image saturation within the trail mask area to obtain the final trail enhanced image.

[0047] Step S4 aims to enhance the color saturation of the 8-bit image obtained in step S3, while using the mask obtained in step S2 to specify the areas to be enhanced. In this step, the image is first converted from RGB to HSV color space. This allows for precise adjustment of hue, saturation, and brightness without affecting other color aspects. This method is chosen because high-resolution images capture richly detailed trails, necessitating meticulous enhancement to maintain image quality and signal-to-noise ratio. Adjusting saturation enhances the trails without affecting other image properties, thus preserving the detail and quality of the high-resolution image. The conversion formula between RGB and HSV is as follows: (11) (12) (13) In this formula, the saturation (S) in the HSV image is enhanced within the region defined by the mask using a predetermined coefficient, which is defined as follows: (14) in, S original This represents the original saturation value of any pixel within the BTR value of the mask, while s >1 is the predetermined coefficient for enhancing saturation. S Enhanced This is the enhanced saturation value.

[0048] Finally, the enhanced saturation and brightness are merged back into the HSV image and then converted back to RGB. The result of this process is enhanced saturation in the whet region, highlighting the whet, while the background remains unaffected.

[0049] To better demonstrate the effectiveness of the wake enhancement method of this invention, we chose to compare it with channel subtraction methods commonly used for low-resolution images, as well as a method that simultaneously increases brightness and saturation within the wake mask. The evaluation metric we chose is the signal-to-noise ratio (SCR), which is calculated as follows: in, G mt It is the average gray value of the wake region. G mb It is the average gray value of the local background area, while This is the standard deviation of the local background region. This metric is crucial because it quantifies the difference between the wake and the surrounding noise, a key factor characterizing the effectiveness of TIR image enhancement.

[0050] The enhanced model of the invention is illustrated using representative images, such as... Figure 7 As shown, images are classified into long ( ) frames against diverse backgrounds including oceans, vegetation, urban areas, and clouds. Figure 7 Medium (b) and short ( Figure 7 (a) The trail in the middle is used to fully showcase the enhanced effect. Figure 7 The enhancement results reveal that the method of this invention achieves the highest signal-to-noise ratio (SCR) for both long and short wakes. Traditional channel subtraction methods yield the lowest SCR, while methods that simultaneously enhance brightness and saturation do improve the SCR of the original image, but none surpass the method of this invention that enhances only saturation.

[0051] The proposed wake enhancement algorithm first crops the original satellite image to obtain a 16-bit smaller image, enabling targeted and precise processing of the wake. Second, by deeply analyzing the physical characteristics and infrared imaging properties of the wake, a mask for wake enhancement is generated, allowing for the selection of the wake enhancement region. Next, the cropped 16-bit smaller image is percentage-truncated and stretched to effectively reduce or eliminate extreme grayscale values ​​typically caused by noise, cloud cover, or image edge effects, which can interfere with the accuracy and efficiency of wake enhancement. Finally, the saturation of the image is extracted and enhanced using RGB to HSV color space conversion, achieving effective wake enhancement while preserving image detail features.

[0052] Compared with existing technologies, this invention demonstrates several innovations and advantages. First, by analyzing the physical principles and applying the brightness temperature characteristics of contrails, and then setting the brightness temperature magnitude of each spectral band and the percentile threshold for the brightness temperature ratio between spectral bands, this invention can accurately identify and enhance aviation contrails in high-resolution thermal infrared images, significantly improving the visibility and accuracy of contrails analysis. Second, in terms of image processing, the percentage-truncation linear stretching and saturation enhancement techniques employed in this invention significantly improve the visual effect of the image, making the contrails more prominent and facilitating further detection and analysis.

[0053] This invention provides a comprehensive method that effectively enhances the detection of aircraft contrails in high-resolution three-band thermal infrared remote sensing images by combining multiple data processing techniques. It not only improves processing efficiency but also significantly enhances the detection accuracy and image quality of contrails in high-resolution three-band thermal infrared images through meticulous mask generation and image enhancement strategies. It has broad application value and development prospects.

[0054] The embodiments described above provide a detailed explanation of the technical solutions and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics, characterized in that, Includes the following steps: (1) Cropping satellite thermal infrared images containing flight contrails into sub-images of uniform size; (2) For each sub-image, generate a trail enhancement mask based on the brightness temperature characteristics of the trail, including: (2-1) For each sub-image, calculate the brightness temperature value of each pixel in the sub-image across multiple spectral bands and the brightness temperature ratio between multiple spectral bands. The formula for calculating the brightness temperature ratio is: in: T 1 represents the brightness temperature of the B1 spectral band. T 2 represents the brightness temperature of the B2 spectral band. T 3 indicates the brightness temperature of the B3 spectral band; R 1. R 2. R 3 represents the ratio of the brightness temperature of the B1 spectral band to the brightness temperature of the B3 spectral band, the ratio of the brightness temperature of the B2 spectral band to the brightness temperature of the B3 spectral band, and the ratio of the brightness temperature of the B1 spectral band to the brightness temperature of the B2 spectral band, respectively. (2-2) For each sub-image, the brightness temperature value and brightness temperature ratio are sorted by magnitude. Based on the brightness temperature magnitude and brightness temperature ratio characteristics of the flight contrail, percentile-based brightness temperature value thresholds and brightness temperature ratio thresholds are selected, including: (2-2i) Sort the brightness temperature values ​​by magnitude: in: For B i The order of brightness temperature values ​​of the spectral bands T i (n) represents B i The brightness temperature values ​​of the spectral bands are sorted by size. n The element at position, n This represents the total number of pixels in the sub-image. (2-2ii) Set the brightness temperature threshold value according to the brightness temperature characteristics of the flight contrail, using the following formula: in: t It is the B that is being filtered. i The percentage position of the brightness temperature values ​​in the spectral bands; TH Ti It is the B that is being filtered. i Brightness temperature threshold for the spectral band; n This represents the total number of pixels in the sub-image. Indicates to t Round down; (2-2iii) Sort the brightness-temperature ratios by magnitude: in: Brightness-to-temperature ratio R j The sorting; R j ( n )for The Middle n The element at the specified position; n This represents the total number of pixels in the sub-image. (2-2iv) Set the brightness temperature ratio threshold based on the brightness temperature characteristics of the flight contrail: in: TH Rj It is a filter. The brightness temperature ratio threshold; r j It is a filter. Percentage position in; (2-3) Define masks based on the selected brightness temperature threshold and brightness temperature ratio threshold, including: (2-3i) Two types of masks are defined based on the brightness temperature threshold and the brightness temperature ratio threshold: The mask for the wake enhancement region is calculated using the following formula. Mask : ； (3) Perform percentage-based truncation and linear stretching on each subgraph; (4) For the stretched image generated in step (3), the image saturation in the area of ​​the trail enhancement mask generated in step (2) is enhanced to obtain the trail enhancement image.

2. The high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics according to claim 1, characterized in that, In step (2-1), the formula for calculating the brightness temperature value is: L i = GAIN i •DN i +BLAS i , i =1,2,3 Where: L i For B i Radiance of the spectral band; h It is Planck's constant. c It's the speed of light. k It is Boltzmann's constant. λ It is the center wavelength. k 1 and k 2 is a constant; T i B i Brightness temperature value of the spectral band; GAIN i For B i The calibration gain coefficient of the spectral band, DN i For B i The effective recorded value of a spectral pixel. BIAS i For B i The calibration offset of the spectral band.

3. The high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics according to claim 1, characterized in that, Step (3) includes: (3-1) Calculate the 0% and 95% percentiles of each pixel in the sub-image; (3-2) Using the pixel values ​​at the 0th and 95th percentiles as the lower and upper limits of stretching, respectively, the pixel values ​​of the sub-image are linearly stretched to generate the stretched image.

4. The high-resolution three-band thermal infrared airborne contrail enhancement method based on brightness temperature characteristics according to claim 1, characterized in that, Step (4) includes: (4-1) Convert the stretched image from an RGB image to an HSV image; (4-2) Based on the trail enhancement mask generated in step (2), the saturation in the HSV image is enhanced using the following formula: in: S original This represents the original saturation value of any pixel within the trail enhancement mask. s >1 is a predetermined coefficient for saturation enhancement. S Enhanced This is the enhanced saturation value; (4-3) Merge the enhanced saturation and brightness back into the HSV image, and then convert it back into an RGB image.

Images