A method and system for measuring atmospheric visibility

Abstract

The present invention relates to a method and system for measuring atmospheric visibility, belonging to the technical field of visibility measurement, and solves the problem of inaccurate visibility measurement accuracy in the prior art. Specifically, it includes: obtaining a first RGB image including the sky background; and performing dark channel extraction on the obtained first RGB image to obtain a first dark channel image; performing atmospheric light value normalization based on the first dark channel image to obtain a second RGB image after atmospheric light value normalization; performing dark image processing on the second RGB image to obtain a second dark channel image; calculating the transmittance of the object in the second dark channel image according to the atmospheric light value and the dark channel value of the second dark channel image; obtaining the visibility value of the second dark channel image based on the transmittance of the object in the second dark channel image and displaying it. It realizes that the calculation process of visibility is not affected by the change of atmospheric light, and improves the measurement accuracy and robustness.

Description

Technical Field

[0001] The present invention relates to the technical field of visibility measurement, and in particular to a method and system for measuring atmospheric visibility. Background Art

[0002] Many fields such as transportation, security, and meteorological monitoring require accurate measurement of atmospheric visibility values. The current visibility measurement equipment and methods mainly include observation method, infrared visibility meter, and image processing method. Among them, the observation method relies on the human eye and the experience of the observer to give an estimated atmospheric visibility value, and its value accuracy is low. The infrared visibility meter uses the principle of forward scattering of light to accurately measure the visibility value, but the infrared visibility meter only detects the visibility value of a certain point and is expensive. The image processing method relies on a common device such as a camera to achieve large-scale visibility monitoring, and the cost is low, which is very suitable for monitoring atmospheric visibility values. At present, the image processing methods for monitoring visibility values ​​mainly include contrast method and brightness difference method. Among them, the contrast method and the brightness difference method both rely on objects with multiple tones. If only objects with a single tone are observed, their accuracy will be reduced. However, in actual applications, scenes with a single tone are mostly, such as a building. Scenes with multiple tones are difficult to find, and the atmospheric illumination is different at different times, which affects the accuracy of visibility calculated by various algorithms. Summary of the invention

[0003] In view of the above analysis, an embodiment of the present invention aims to provide a method for automatic normalization of atmospheric light intensity and visibility measurement, so as to solve the problem of inaccurate visibility measurement accuracy.

[0004] On the one hand, an embodiment of the present invention provides a method for measuring atmospheric visibility, comprising:

[0005] Acquire a first RGB image including a sky background;

[0006] Perform dark channel extraction according to the acquired first RGB image to obtain a first dark channel image;

[0007] Normalizing the atmospheric illumination value according to the first dark channel image to obtain a second RGB image after the atmospheric illumination value is normalized;

[0008] Performing dark image processing according to the second RGB image to obtain a second dark channel image;

[0009] Calculating the transmittance of the object in the second dark channel image according to the atmospheric illumination value and the dark channel value of the second dark channel image;

[0010] Based on the transmittance of the object in the second dark channel image, a visibility value of the second dark channel image is obtained and displayed.

[0011] Optionally, performing dark channel extraction according to the acquired first RGB image to obtain a first dark channel image includes:

[0012] Acquire a first RGB image including a sky background through a camera; the camera is a fixed-focus camera or a zoom camera; the fixed-focus camera or the zoom camera includes: a visible light camera or an infrared camera;

[0013] The camera includes a photosensitive element, which is a CCD, a CMOS or an avalanche diode; when the camera is shooting, if the atmospheric light intensity is insufficient, laser, infrared light or illumination light is used for supplementary light;

[0014] Select the minimum color value of each pixel in the first RGB image as the dark channel value;

[0015] The first dark channel image is obtained by combining all dark channel values ​​of the entire image.

[0016] Optionally, the normalizing the atmospheric illumination value according to the first dark channel image to obtain the second RGB image after the atmospheric illumination value is normalized includes:

[0017] Acquire the atmospheric illumination value according to the first dark channel image containing the sky background;

[0018] Adjust the camera based on the comparison between the acquired atmospheric illumination value and the standard value;

[0019] The adjusting of the camera includes: adjusting the exposure, adjusting the aperture or other parameters that affect the image brightness; the adjusting of the exposure includes: adjusting by using a PID adjustment algorithm, an automatic control algorithm or a non-automatic adjustment algorithm;

[0020] The adjusted camera is used to obtain a second RGB image after normalization of the atmospheric illumination value.

[0021] Optionally, comparing the atmospheric illumination value with a standard value includes:

[0022] If the atmospheric illumination is higher than the standard value, the exposure adjustment value is negative;

[0023] If the atmospheric illumination is lower than the standard value, the exposure adjustment value is positive.

[0024] Optionally, calculating the transmittance of the object in the second dark channel image according to the atmospheric illumination value and the dark channel value of the second dark channel image includes:

[0025] Remove the sky background pixels in the second dark channel image;

[0026] Using image segmentation method to segment different objects in the second dark channel image;

[0027] One or more objects are selected as research objects, and the transmittance of the research objects is obtained by calculation; the distance between the research object and the camera is obtained by camera matrix measurement, and a transmittance fitting curve is obtained according to the distance measured by the camera matrix; the transmittance of the research object is obtained based on the transmittance fitting curve.

[0028] Optionally, selecting one or more objects as research objects and obtaining the transmittance of the research objects by calculation includes:

[0029] Calculate the transmittance corresponding to each object based on the obtained dark channel value and atmospheric illumination value:

[0030]

[0031] Among them, t C is the transmittance value corresponding to the object, A is the atmospheric illumination value, ω is the compensation coefficient, I c is the dark channel value.

[0032] Optionally, obtaining the visibility value of the second dark channel image based on the transmittance of the object in the second dark channel image includes:

[0033] Calculating a fitting relationship according to the transmittance of the object in the second dark channel image;

[0034] According to the calculated fitting relationship, the visibility value of the second dark channel image is obtained.

[0035] Optionally, when the selected research object is an object, obtaining the visibility value of the second dark channel image includes:

[0036] The visibility value of the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0037] V=a*t c +b

[0038] Wherein, a and b are fitting parameters obtained by data fitting, and V is the visibility value of the obtained second dark channel image;

[0039] When the selected research object is a plurality of objects, the visibility value of the second dark channel image is obtained, including:

[0040] The visibility value of the area corresponding to each object in the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0041] V i =a*t i +b

[0042] Among them, a and b are fitting parameters, Vi is the visibility value of the area corresponding to the i-th object, and ti is the transmittance of the i-th object;

[0043] The visibility value V of the second dark channel image is obtained by averaging the visibility values ​​Vi of the areas corresponding to all objects.

[0044] An embodiment of the present invention provides another atmospheric visibility measurement system, comprising: a camera, a processor, and a display;

[0045] The camera is used to obtain a first RGB image including a sky background image and having a resolution of 1440*1080, and a second RGB image with a constant atmospheric illumination value; the processor is used to process the obtained first / second RGB image to obtain a visibility value;

[0046] A display is used to display the final visibility value obtained.

[0047] Optionally, the processor includes: a dark channel extraction module, an atmospheric illumination module, and a visibility calculation module;

[0048] The dark channel extraction module is used to perform dark channel extraction on the acquired first RGB image to obtain a first dark channel image; and to perform dark channel extraction on the second RGB image to obtain a second dark channel image;

[0049] The atmospheric illumination module is used to obtain an atmospheric illumination value according to the obtained first dark channel image; the atmospheric illumination value is used to adjust the exposure of the camera to obtain a second RGB image with a constant atmospheric illumination value;

[0050] The visibility calculation module is used to calculate the transmittance of the second dark channel image according to the second dark channel value and the atmospheric illumination value, and then obtain the visibility of the second dark channel image according to the transmittance of the second dark channel image.

[0051] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0052] 1. The exposure is adjusted to achieve a constant atmospheric illumination value of the image and complete the image normalization operation. Under a constant atmospheric illumination value, the transmittance is calculated by combining the image dark channel and the atmospheric propagation model. The visibility value is calculated based on the transmittance combined with the prior fitting curve.

[0053] 2. The dark channel does not depend on the color diversity of the target object, so its application scenarios are wider. The constant atmospheric light value ensures that the visibility calculation process will not be affected by the change of atmospheric illumination, which improves the measurement accuracy and robustness. Therefore, the present invention has the advantages of wide application range and measurement stability.

[0054] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

[0056] Figure 1 This is a flow chart of a method for measuring atmospheric visibility in an embodiment of the present invention;

[0057] Figure 2 This is a schematic diagram of a dark channel image before processing in an embodiment of the present invention;

[0058] Figure 3 Schematic diagram of dark channel image processing in an embodiment of the present invention

[0059] Figure 4 This is a flow chart of normalizing atmospheric illumination values ​​in an embodiment of the present invention;

[0060] Figure 5 This is a flow chart of comparing the atmospheric illumination value with the standard value in an embodiment of the present invention;

[0061] Figure 6 A system structure diagram of atmospheric visibility measurement in an embodiment of the present invention;

[0062] Figure 7 FIG. 4 is a diagram showing the structure of a processor in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0064] A specific embodiment of the present invention discloses a method for measuring atmospheric visibility, such as Figure 1 Specifically including:

[0065] Step S1: Acquire a first RGB image including a sky background.

[0066] Exemplarily, an image including a sky background may be captured by an RGB camera with a resolution of 1440*1080 to obtain a first RGB image, which is then transmitted to a processing device for subsequent processing steps, wherein the processing device may be a host computer, etc.

[0067] Step S2: Perform dark channel extraction on the obtained first RGB image to obtain a first dark channel image.

[0068] In an RGB image, each pixel has three color values of RGB. The minimum value of these three color values for each pixel is taken as the dark channel value, and all the dark channel values of the image are combined to form a new image. The new image obtained at this time is the dark channel image. It can be expressed by the formula:

[0069] I C = min(I R , I G , I B )

[0070] where I c is the dark channel value, I R is the pixel value of red, I G is the pixel value of green, and I B is the pixel value of blue.

[0071] It should be noted that in the case of no fog, the dark channel value of the image is almost 0. In the case of fog and very thick fog, the dark channel value approaches the atmospheric light value. From this, it can be known that there is a certain correlation between the dark channel and the fog concentration. As Figure 2-3 shown.

[0072] Obtain a first RGB image including the sky background through a camera; the camera is a fixed-focus camera or a zoom camera; the fixed-focus camera or zoom camera includes: a visible light camera or an infrared camera;

[0073] The camera includes a photosensitive element, and the photosensitive element is a CCD, a CMOS or an avalanche diode; when the camera takes pictures, if the atmospheric light intensity is insufficient, laser, infrared light or illumination light is used for fill light.

[0074] Step S3: Perform atmospheric light value normalization based on the first dark channel image to obtain a second RGB image with the atmospheric light value normalized.

[0075] Specifically, since the atmospheric illumination value is a measure of the ambient light in which the image is taken, the ambient light has a great influence on optical analysis and calculation. In the captured image of the embodiment of the present invention, there must be a sky background in the image. Since the sky background is not contaminated by reflected light, and the light of the sky background is all scattered light, which is close to the ambient light, the atmospheric illumination value has a great influence on the visibility of the image, which seriously affects the accuracy of visibility measurement. Therefore, the atmospheric illumination value of the image needs to be kept constant to improve the accuracy of visibility measurement. The embodiment of the present invention obtains the atmospheric illumination value of the image, thereby adjusting the camera exposure or aperture to keep the atmospheric illumination value constant. This then achieves the normalization operation of the atmospheric illumination values ​​of all images. Figure 4 shown.

[0076] The specific process is: if the atmospheric light is higher than the standard value, it means that the atmospheric light is strong, and the exposure needs to be reduced to maintain stability, and the exposure adjustment value is negative; if the atmospheric light is lower than the standard value, it means that the atmospheric light is weak, and the exposure needs to be increased to maintain stability, and the exposure adjustment value is positive. Figure 4 shown.

[0077] It should be noted that: if the atmospheric illumination value obtained by the atmospheric illumination module is within the range of 214 to 216, no adjustment is required. If the atmospheric illumination value is not within this range, the PID adjustment method is used, with 215 as the adjustment target value, and the atmospheric illumination value is adjusted to be within the range of 212 to 218 by adjusting the exposure.

[0078] The standard value is the atmospheric illumination value obtained through multiple experiments, ranging from 210 to 220. The atmospheric illumination value within this range has the least interference on image visibility.

[0079] The RGB camera is adjusted according to the obtained atmospheric illumination deviation, and the second RGB image is obtained by shooting with the adjusted camera. The second RGB image obtained at this time is the RGB image normalized by the atmospheric illumination value.

[0080] The adjusting of the camera includes: adjusting the exposure, adjusting the aperture or other parameters that affect the image brightness; the adjusting of the exposure includes: adjusting by using a PID adjustment algorithm, an automatic control algorithm or a non-automatic adjustment algorithm;

[0081] The adjusted camera is used to obtain a second RGB image after normalization of the atmospheric illumination value.

[0082] Step S4: performing dark image processing on the second RGB image to obtain a second dark image;

[0083] Specifically, the processing method is the same as step S2.

[0084] Step S5: calculating the transmittance of the second dark image according to the atmospheric illumination value and the dark channel value of the second dark image, thereby obtaining the visibility of the image;

[0085] Specifically, it includes:

[0086] Step S51, removing sky background pixels in the second dark channel image; specifically, in the embodiment of the present invention, the pixel value of the sky background part in the dark channel image is the atmospheric illumination value.

[0087] Since the transmittance value of the sky background is useless for visibility calculation, the sky background pixels need to be removed first.

[0088] Step S52, using an image segmentation method to segment different objects in the second dark channel image;

[0089] Specifically, an image segmentation method is used to segment different objects contained in the image.

[0090] Step S53: select one or more objects as research objects, and obtain the transmittance of the research objects by calculation.

[0091] The distance between the research object and the camera is obtained by using a camera matrix measurement, and a transmittance fitting curve is obtained according to the distance measured by the camera matrix; and the transmittance of the research object is obtained based on the transmittance fitting curve.

[0092] Among them, the transmittance corresponding to each object is calculated according to the obtained dark channel value and atmospheric illumination value, and the calculation formula is:

[0093]

[0094] Among them, t C is the transmittance value corresponding to the object, A is the atmospheric illumination value, ω is the compensation coefficient, I c is the dark channel value. The above formula can be used to calculate the transmittance value corresponding to each pixel of the dark channel image. The larger the transmittance value, the higher the visibility.

[0095] Step S54: According to the transmittance of the selected research object, the transmittance of the research object is obtained based on the transmittance fitting curve, and then the visibility is calculated.

[0096] Specifically, after verifying that there is a linear relationship between transmittance and visibility, a large amount of experimental data is used to fit the linear relationship to obtain a fitting function.

[0097] When the selected research object is an object, the visibility value of the second dark channel image is obtained including:

[0098] The visibility value of the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0099] V=a*t 均 +b

[0100] Among them, a and b are fitting parameters obtained by data fitting, and V is the visibility value. Thus, the visibility value V can be calculated.

[0101] In the embodiment of the present invention, one or more objects may be selected as research objects.

[0102] Specifically, if an object is selected as the research object, there may be slightly different appearances of the same object, resulting in slightly different transmittances. The transmittance of the research object can be obtained and then the average value can be calculated to accurately calculate the transmittance. If the difference is not large, it can be ignored, and the transmittance value of the research object can be used as the transmittance of the dark image.

[0103] If multiple objects are selected as research objects, the transmittance of each research object must be obtained. Each transmittance is calculated through the fitting relationship to obtain a visibility value, and then the visibility values ​​are averaged to obtain an average visibility value, which is the visibility of the second dark image.

[0104] Specifically, when the selected research object is a plurality of objects, the visibility value of the second dark channel image is obtained, including:

[0105] The visibility value of the area corresponding to each object in the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0106] V i =a*t i +b

[0107] Among them, a and b are fitting parameters, V i is the visibility value of the area corresponding to the i-th object, t i is the transmittance of the ith object;

[0108] The visibility value V for the area corresponding to all objects i The visibility value V of the second dark channel image is obtained by averaging.

[0109] Step S6: The display shows the visibility value of the dark channel image;

[0110] Specifically, the visibility value V of the second dark channel image finally obtained by calculation is displayed on the display.

[0111] Another embodiment of the present invention discloses a system for measuring atmospheric visibility, such as Figure 5 As shown, specifically including:

[0112] The camera is used to obtain a first RGB image including a sky background image and having a resolution of 1440*1080, and to obtain a second RGB image with a constant atmospheric illumination value.

[0113] The obtained first RGB image is transmitted to a processor, wherein the processor may be a host computer, etc.

[0114] Acquire a first RGB image including a sky background through a camera; the camera is a fixed-focus camera or a zoom camera; the fixed-focus camera or the zoom camera includes: a visible light camera or an infrared camera;

[0115] The camera includes a photosensitive element, which is a CCD, a CMOS or an avalanche diode. When the camera is shooting, if the atmospheric light intensity is insufficient, laser, infrared light or illumination light is used for fill light.

[0116] Based on the comparison between the atmospheric illumination value and the standard value, the exposure is adjusted to achieve a constant atmospheric illumination value.

[0117] Specifically, the atmospheric illumination value is obtained through the atmospheric illumination module in the processor, and the atmospheric illumination value is compared with the standard value to obtain the atmospheric illumination deviation of the current image. The atmospheric illumination deviation is calculated according to the PID adjustment method to calculate the corresponding exposure adjustment value. Generally, if the atmospheric illumination is higher than the standard value, it means that the atmospheric illumination is strong, and the exposure needs to be reduced to maintain stability, and the exposure adjustment value is negative; if the atmospheric illumination is lower than the standard value, it means that the atmospheric illumination is weak, and the exposure needs to be increased to maintain stability, and the exposure adjustment value is positive.

[0118] It should be noted that: if the atmospheric illumination value obtained by the atmospheric illumination module is within the range of 214 to 216, no adjustment is required. If the atmospheric illumination value is not within this range, the PID adjustment method is used, with 215 as the adjustment target value, and the atmospheric illumination value is adjusted to be within the range of 212 to 218 by adjusting the exposure.

[0119] The standard value is the atmospheric illumination value obtained through multiple experiments, ranging from 210 to 220. The atmospheric illumination value within this range has the least interference on image visibility.

[0120] The exposure of the RGB camera is adjusted according to the obtained atmospheric illumination deviation, and the second RGB image is obtained by shooting with the adjusted camera. The second RGB image obtained at this time is the RGB image normalized by the atmospheric illumination value.

[0121] The adjusting of the camera includes: adjusting the exposure, adjusting the aperture or other parameters that affect the image brightness; the adjusting of the exposure includes: adjusting by using a PID adjustment algorithm, an automatic control algorithm or a non-automatic adjustment algorithm;

[0122] The adjusted camera is used to obtain a second RGB image after normalization of the atmospheric illumination value.

[0123] The specific adjustment process is as follows: Figure 2 shown.

[0124] The second RGB image is obtained by shooting with the adjusted camera.

[0125] The processor is used to process the obtained first / second RGB image; it includes a dark channel extraction module, an atmospheric illumination module, and a visibility calculation module.

[0126] The dark channel extraction module performs dark channel extraction on the acquired first RGB image to obtain a first dark channel image; and is used to perform dark channel extraction on the second RGB image to obtain a second dark channel image.

[0127] Specifically, in the first RGB image obtained, each pixel has three RGB color values, the minimum value of these three values ​​is taken as the dark channel value, and all the image dark channel values ​​are combined into a new image, which is the first dark channel image. Figure 3 shown.

[0128] The new image obtained by combining all the dark channel values ​​of the image can be expressed as:

[0129] I C =min(I R ,I G ,I B )

[0130] Among them, I c is the dark channel value, I R is the red pixel value, I G is the green pixel value, I B The pixel value is blue.

[0131] In the absence of fog, the value of the dark channel of the image is almost 0. In the presence of fog and very heavy fog, the value of the dark channel approaches the atmospheric illumination value. It can be seen that the dark channel has a correlation with the fog concentration. In the embodiment of the present invention, the fog concentration represents visibility.

[0132] The atmospheric illumination module is used to perform normalization processing on the atmospheric illumination according to the obtained first dark channel image.

[0133] Specifically, since the atmospheric illumination value has a great influence on the visibility extraction of the image, which seriously affects the visibility measurement accuracy, the atmospheric illumination value of the image needs to be kept constant to improve the measurement accuracy. The embodiment of the present invention calculates the atmospheric illumination value of the image, and adjusts the camera exposure or aperture according to the atmospheric illumination value of the image to keep the atmospheric illumination value constant. This realizes the normalization operation of the atmospheric illumination values ​​of all images. Figure 4 shown.

[0134] The visibility calculation module is used to obtain the transmittance of the second dark image by calculation according to the second dark channel value and the atmospheric illumination value, and then obtain the visibility of the second dark image according to the transmittance of the second dark channel image.

[0135] Specifically, the sky background pixels in the second dark channel image are removed; specifically, in the embodiment of the present invention, the pixel value of the sky background part in the dark channel image is the atmospheric illumination value.

[0136] Since the transmittance value of the sky background is useless for visibility calculation, the sky background pixels need to be removed first.

[0137] Secondly, the image segmentation method is used to segment different objects in the second dark channel image;

[0138] Specifically, an image segmentation method is used to segment different objects contained in the image.

[0139] Next, one or more objects are selected as research objects, and the transmittance of the research objects is obtained by calculation.

[0140] Among them, the transmittance corresponding to each object is calculated according to the obtained dark channel value and atmospheric illumination value, and the calculation formula is:

[0141]

[0142] Among them, t C is the transmittance value corresponding to the object, A is the atmospheric illumination value, ω is the compensation coefficient, I c is the dark channel value. The above formula can be used to calculate the transmittance value corresponding to each pixel of the dark channel image. The larger the transmittance value, the higher the visibility.

[0143] According to the transmittance of the selected research object, the transmittance of the research object is obtained based on the transmittance fitting curve, and then the visibility is calculated.

[0144] Specifically, after verifying that there is a linear relationship between average transmittance and visibility, a large amount of experimental data is used to fit the linear relationship to obtain a fitting function.

[0145] When the selected research object is an object, the visibility value of the second dark channel image is obtained including:

[0146] The visibility value of the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0147] V=a*t 均 +b

[0148] Among them, a and b are fitting parameters obtained by data fitting. From this, the visibility value V can be calculated.

[0149] In the embodiment of the present invention, one or more objects may be selected as research objects.

[0150] Specifically, if an object is selected as the research object, there may be slightly different appearances of the same object, resulting in slightly different transmittances. The transmittance of the research object can be obtained and then the average value can be calculated to accurately calculate the transmittance. If the difference is not large, it can be ignored, and the transmittance value of the research object can be used as the transmittance of the dark image.

[0151] If multiple objects are selected as research objects, the transmittance of each research object must be obtained. Each transmittance is calculated through the fitting relationship to obtain a visibility value, and then the visibility values ​​are averaged to obtain an average visibility value, which is the visibility value of the second dark image.

[0152] Specifically, when the selected research object is a plurality of objects, the visibility value of the second dark channel image is obtained, including:

[0153] The visibility value of the area corresponding to each object in the second dark channel image is obtained based on the fitting function between the following transmittance and visibility:

[0154] V i =a*t i +b

[0155] Among them, a and b are fitting parameters, V i is the visibility value of the area corresponding to the i-th object, t i is the transmittance of the ith object;

[0156] The visibility value V for the area corresponding to all objects i The visibility value V of the second dark channel image is obtained by averaging.

[0157] Finally, the visibility value V obtained by calculation is displayed on the display.

[0158] In addition, the atmospheric illumination value can be adjusted by adjusting the camera exposure, or the camera aperture or other parameters that can adjust the amount of illumination received by the image can be adjusted to achieve automatic adjustment.

[0159] The present invention is based on the automatic normalization of atmospheric light intensity and the visibility measurement method, and adopts the method of adjusting the exposure to achieve the constant atmospheric light value, thereby completing image normalization. Under the constant atmospheric light value, the visibility value is calculated by combining the dark channel value and the transmittance of atmospheric transmission, and the result is obtained from the display.

[0160] The present invention proposes a method for automatic normalization of atmospheric illumination intensity and visibility measurement, which uses the exposure adjustment method to achieve a constant atmospheric illumination value of an image, completes the image normalization operation, and calculates the transmittance by combining the dark channel value of the image with the atmospheric propagation model under a constant atmospheric illumination value, and calculates the visibility value based on the transmittance combined with the prior fitting curve. The dark channel does not depend on the color diversity of the target object, so its application scenarios are more extensive, and the constant atmospheric light value ensures that the visibility calculation process will not be affected by changes in atmospheric illumination, thereby improving the measurement accuracy and robustness.

[0161] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0162] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for measuring atmospheric visibility, characterized in that, comprising: obtaining a first RGB image including the sky background through a camera; performing dark channel extraction on the obtained first RGB image to obtain a first dark channel image, including: selecting the minimum value of each pixel color in the first RGB image as the dark channel value; combining all the dark channel values of the entire image to obtain the first dark channel image; performing atmospheric light value normalization on the first dark channel image to obtain a second RGB image after atmospheric light value normalization, including: obtaining the atmospheric light value according to the first dark channel image containing the sky background; comparing the obtained atmospheric light value with a standard value and adjusting the camera; using the adjusted camera to capture a second RGB image after atmospheric light value normalization; performing dark image processing on the second RGB image to obtain a second dark channel image; calculating the transmittance of an object in the second dark channel image according to the atmospheric light value and the dark channel value of the second dark channel image, including: removing the sky background pixels in the second dark channel image; segmenting different objects in the second dark channel image by using an image segmentation method; selecting one or more objects as the research object and calculating the transmittance of the research object; the distance between the research object and the camera is measured by a camera matrix, and a transmittance fitting curve is obtained according to the distance measured by the camera matrix; obtaining the transmittance of the research object based on the transmittance fitting curve; obtaining the visibility value of the second dark channel image based on the transmittance of the object in the second dark channel image and displaying it.

2. The method for measuring atmospheric visibility according to claim 1, characterized in that, the camera is a fixed-focus camera or a zoom camera; the fixed-focus camera or zoom camera includes: a visible light camera or an infrared camera; the camera includes a photosensitive element, and the photosensitive element is a CCD, a CMOS or an avalanche diode; when the camera captures, if the atmospheric light intensity is insufficient, laser, infrared light or illumination light is used for supplementary lighting.

3. The method for measuring atmospheric visibility according to claim 1, characterized in that, the adjustment of the camera includes: adjusting the exposure amount, adjusting the aperture or other parameters affecting the image brightness; the adjustment of the exposure amount includes: adjusting by using a PID adjustment algorithm, an automatic control algorithm or a non-automatic adjustment algorithm.

4. The method for measuring atmospheric visibility according to claim 3, characterized in that, the comparison of the atmospheric light value with the standard value includes: if the atmospheric light is higher than the standard value, the exposure amount adjustment value is negative; if the atmospheric light is lower than the standard value, the exposure amount adjustment value is positive.

5. The method for measuring atmospheric visibility according to claim 1, characterized in that, the selection of one or more objects as the research object and the calculation of the transmittance of the research object include: calculating the transmittance corresponding to each object according to the dark channel value and the atmospheric light value of the obtained second dark channel image: where t C is the transmittance value corresponding to the object, A is the atmospheric light value, ω is the compensation coefficient, and I c is the dark channel value of the second dark channel image.

6. The method for measuring atmospheric visibility according to claim 1, characterized in that, Obtaining the visibility value of the second dark channel image based on the transmittance of the object in the second dark channel image includes: Performing fitting relationship calculation according to the transmittance of the object in the second dark channel image; Obtaining the visibility value of the second dark channel image according to the calculated fitting relationship.

7. The method for measuring atmospheric visibility according to claim 6, wherein, when the selected research object is an object, obtaining the visibility value of the second dark channel image includes: Obtaining the visibility value of the second dark channel image based on the following fitting function between transmittance and visibility: V = a * t c + b where a and b are fitting parameters obtained by data fitting, and V is the visibility value of the obtained second dark channel image; when the selected research object is multiple objects, obtaining the visibility value of the second dark channel image includes: Obtaining the visibility value of the region corresponding to each object in the second dark channel image based on the following fitting function between transmittance and visibility: V i = a * t i + b where a and b are fitting parameters, V i is the visibility value of the area corresponding to the i-th object, and t i is the transmittance of the i-th object; Visibility value V of the regions corresponding to all objects i Average to obtain the visibility value V of the second dark channel image.

8. An atmospheric visibility measurement system, adopting the method for measuring atmospheric visibility according to any one of claims 1-7, wherein, it includes: a camera, a processor, and a display; The camera is used to obtain a first RGB image including a sky background image with a resolution of 1440*1080, and is also used to obtain a second RGB image after normalizing the atmospheric light value; the processor is used to process the obtained first / second RGB image to obtain the visibility value; The display is used to display the obtained final visibility value.

9. The atmospheric visibility measurement system according to claim 8, wherein, the processor includes: a dark channel extraction module, an atmospheric light module, and a visibility calculation module; The dark channel extraction module is used to perform dark channel extraction on the obtained first RGB image to obtain a first dark channel image; and is also used to perform dark channel extraction on the second RGB image to obtain a second dark channel image; The atmospheric light module is used to obtain the atmospheric light value according to the obtained first dark channel image; the atmospheric light value is used to adjust the exposure of the camera to obtain a second RGB image after normalizing the atmospheric light value; The visibility calculation module is used to calculate the transmittance of the second dark channel image according to the dark channel value and the atmospheric light value of the second dark channel image, and then obtain the visibility of the second dark channel image according to the transmittance of the second dark channel image.

Images