Low-illumination image enhancement method and device, and image processing equipment

Abstract

The application provides a low-illumination image enhancement method, device and image processing equipment, which comprises the following steps: acquiring an original image and determining the normalized pixel value of each pixel point in the original image; acquiring a pixel block centered on each pixel point and determining the pixel point parameter of each pixel point and the pixel block parameter of each pixel block based on the normalized pixel value; determining the adaptive gain coefficient and the adaptive correction value of each pixel point according to the pixel point parameter and the pixel block parameter; determining the corrected pixel value of each pixel point according to the adaptive gain coefficient and the adaptive correction value, and obtaining an enhanced image according to the corrected pixel value of each pixel point. The application realizes the purpose of adjusting the pixel value of each pixel point in a low-illumination image based on the adaptive gain coefficient and the adaptive correction value, thereby improving the image quality of the enhanced image.

Description

Technical Field

[0001] This invention relates to the field of image enhancement technology, and specifically to a low-light image enhancement method, apparatus, and image processing equipment. Background Technology

[0002] In most vision systems, low illumination is a common phenomenon due to factors such as light obstruction, differences in material reflectivity, and nighttime shooting scenarios. These phenomena result in low image quality of the scene captured by the camera, manifested as very weak local signals, dark images, poor signal-to-noise ratio, and blurriness. Therefore, to improve the visual effect of images acquired under low illumination, it is necessary to process the image of the subject under low illumination.

[0003] Existing methods for enhancing low-light images include HSV-based gamma transform, deep learning, histogram transform, wavelet transform, and Retinex theory. However, these methods all apply the same enhancement to every pixel in the original low-light image. Because low-light images have a small grayscale range and high spatial correlation between adjacent pixels, details, colors, background, and noise are all contained within a very small grayscale range. While existing enhancement methods can improve the brightness of low-light images, they also lead to problems such as amplified image noise, overexposure of high-light pixel areas, decreased contrast, and loss of detail.

[0004] Therefore, there is an urgent need to provide a low-light image enhancement method, apparatus, and image processing device to reduce image noise while improving the brightness of low-light images, thus ensuring the contrast and detail features of the images. Summary of the Invention

[0005] In view of this, it is necessary to provide a low-light image enhancement method, apparatus and image processing device to solve the technical problems existing in the prior art that, while improving the brightness of low-light images, image noise is amplified, high-light pixel areas are overexposed, contrast deteriorates and detail features cannot be reflected.

[0006] On one hand, the present invention provides a low-light image enhancement method, comprising:

[0007] Acquire the original image and determine the normalized pixel value of each pixel in the original image;

[0008] Obtain pixel blocks centered on each of the aforementioned pixels, and determine the pixel point parameters of the pixels and the pixel block parameters of the pixel blocks based on the normalized pixel values;

[0009] The adaptive gain coefficient and adaptive correction value of the pixel are determined based on the pixel point parameters and the pixel block parameters;

[0010] The corrected pixel value of the pixel is determined based on the adaptive gain coefficient and the adaptive correction value, and the enhanced image is obtained based on the corrected pixel value of the pixel.

[0011] In some possible implementations, determining the normalized pixel value of each pixel in the original image includes:

[0012] Obtain the original red component pixel value, original green component pixel value, original blue component pixel value, and red component normalization coefficient, green component normalization coefficient, and blue component normalization coefficient of each pixel in the original image;

[0013] The normalized pixel value of each pixel is determined based on the original red component pixel value, the original green component pixel value, the original blue component pixel value, the red component normalization coefficient, the green component normalization coefficient, and the blue component normalization coefficient.

[0014] In some possible implementations, the pixel parameters include the pixel transform value after gamma transformation of the pixel value and the pixel transform derivative value; the pixel block parameters include the size of the pixel block, the sum of normalized pixel values, the pixel block mean, and the pixel block mean square error.

[0015] In some possible implementations, determining the adaptive gain coefficient and adaptive correction value of the pixel based on the pixel point parameters and the pixel block parameters includes:

[0016] A gain coefficient segmentation determination model is constructed based on the pixel point parameters and the pixel block parameters, and the adaptive gain coefficient of the pixel point is determined based on the gain coefficient segmentation determination model.

[0017] Determine the maximum adaptive gain coefficient among the adaptive gain coefficients of each pixel;

[0018] The adaptive correction value of the pixel is determined by the model based on the adaptive gain coefficient, the maximum adaptive gain coefficient, and the preset correction value.

[0019] In some possible implementations, the gain coefficient segmentation determination model includes a first gain coefficient determination sub-model, a second gain coefficient determination sub-model, and a third gain coefficient determination sub-model; determining the adaptive gain coefficient of the pixel based on the gain coefficient segmentation determination model includes:

[0020] When the mean value of the pixel block is greater than or equal to the preset lower limit value, or the normalized pixel value is greater than the preset lower limit value and less than or equal to the preset upper limit value, the adaptive gain coefficient of the pixel is determined based on the first gain coefficient determination sub-model.

[0021] When the average value of the pixel block is less than the preset lower limit value, and the normalized pixel value is greater than the average value of the pixel block but less than or equal to the preset lower limit value, the adaptive gain coefficient of the pixel is determined based on the second gain coefficient determination sub-model.

[0022] When the average value of the pixel block is less than the preset lower limit value, and the normalized pixel value is less than or equal to the average value of the pixel block, the adaptive gain coefficient of the pixel is determined based on the third gain coefficient determination sub-model.

[0023] In some possible implementations, determining the corrected pixel value of the pixel based on the adaptive gain coefficient and the adaptive correction value includes:

[0024] The red component correction pixel value of the pixel is determined based on the original red component pixel value, the adaptive gain coefficient, and the adaptive correction value;

[0025] The green component corrected pixel value of the pixel is determined based on the original green component pixel value, the adaptive gain coefficient, and the adaptive correction value;

[0026] The blue component correction pixel value of the pixel is determined based on the original blue component pixel value, the adaptive gain coefficient, and the adaptive correction value.

[0027] In some possible implementations, the low-light image enhancement method further includes:

[0028] Obtain the original pixel block centered on each pixel in the enhanced image, determine the set of adaptive gain coefficients corresponding to the original pixel block, and determine whether there is at least one adaptive gain coefficient in the set of adaptive gain coefficients that is greater than the gain coefficient threshold.

[0029] If all adaptive gain coefficients in the set of adaptive gain coefficients are less than or equal to the gain coefficient threshold, then a target pixel block is determined with each pixel in the enhanced image as the center, and the target pixel block is filtered. The center pixel of the filtered target pixel block replaces the center pixel of the original pixel block to obtain an optimized image.

[0030] If at least one adaptive gain coefficient in the set of adaptive gain coefficients is greater than the gain coefficient threshold, then a target pixel block is determined with each pixel in the enhanced image as the center, and edge enhancement processing is performed on the target pixel block. The center pixel of the edge-enhanced target pixel block replaces the center pixel of the original pixel block to obtain an optimized image.

[0031] In some possible implementations, the low-light image enhancement method further includes:

[0032] The enhanced image and / or the optimized image are subjected to saturation correction processing to obtain a saturation correction image;

[0033] The saturation correction image is subjected to pseudo-color correction processing to obtain a color correction image.

[0034] On the other hand, the present invention also provides a low-light image enhancement device, comprising:

[0035] A normalized pixel value determination unit is used to acquire the original image and determine the normalized pixel value of each pixel in the original image.

[0036] A parameter determination unit is used to obtain a pixel block centered on each pixel point, and to determine the pixel point parameter of the pixel point and the pixel block parameter of the pixel block based on the normalized pixel value;

[0037] A gain coefficient and correction value determination unit is used to determine the adaptive gain coefficient and adaptive correction value of the pixel based on the pixel point parameters and the pixel block parameters;

[0038] An enhanced image determination unit is used to determine the corrected pixel value of the pixel based on the adaptive gain coefficient and the adaptive correction value, and to obtain an enhanced image based on the corrected pixel value of the pixel.

[0039] On the other hand, the present invention also provides an image processing device, including a memory and a processor, wherein,

[0040] The memory is used to store programs;

[0041] The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the low-light image enhancement method described in any of the above implementations.

[0042] The beneficial effects of the above embodiments are as follows: The low-light image enhancement method provided by the present invention determines the adaptive gain coefficient and adaptive correction value of each pixel based on the pixel point parameters and pixel block parameters, and determines the corrected pixel value of the pixel based on the adaptive gain coefficient and adaptive correction value, and obtains the enhanced image based on the corrected pixel value of the pixel. This achieves the purpose of adjusting the pixel value of each pixel in the low-light image based on the adaptive gain coefficient and adaptive correction value, thereby avoiding the technical problems of amplified image noise, overexposure in high-illuminance areas, low contrast, and inability to reflect detailed features caused by uniform adjustment of each pixel in the low-light image, thus improving the image quality of the enhanced image. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A schematic flowchart of an embodiment of the low-light image enhancement method provided by the present invention;

[0045] Figure 2 For the present invention Figure 1 A schematic diagram of an embodiment of S101;

[0046] Figure 3 For the present invention Figure 1 A schematic diagram of an embodiment of S103;

[0047] Figure 4 For the present invention Figure 3 A schematic diagram of an embodiment of S301;

[0048] Figure 5 For the present invention Figure 1 A schematic diagram of an embodiment of S104;

[0049] Figure 6 A schematic diagram of an embodiment of the present invention for obtaining an optimized image;

[0050] Figure 7 This is a schematic diagram of an embodiment of the process for obtaining a color correction image provided by the present invention;

[0051] Figure 8 This is a comparison diagram of image enhancement results provided by the present invention with those of existing technologies;

[0052] Figure 9A schematic diagram of an embodiment of the low-light image enhancement device provided by the present invention;

[0053] Figure 10 This is a schematic diagram of an embodiment of the image processing device provided by the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0055] It should be understood that the illustrative drawings are not drawn to scale. The flowcharts used in this invention illustrate operations implemented according to some embodiments of the invention. It should be understood that the operations in the flowcharts may be implemented out of order, and steps without logical contextual relationships may be reversed or performed simultaneously. Furthermore, those skilled in the art, guided by the content of this invention, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor systems and / or microcontroller systems.

[0056] In the description of the embodiments of the present invention, unless otherwise stated, "and / or" describes the relationship between associated objects, indicating that there can be three relationships, for example: A and / or B, which can represent three cases: A exists alone, A and B exist simultaneously, and B exists alone.

[0057] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0058] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0059] This invention provides a low-light image enhancement method, apparatus, and image processing device, which are described below.

[0060] Figure 1 A schematic flowchart of an embodiment of the low-light image enhancement method provided by the present invention is shown below. Figure 1 As shown, low-light image enhancement methods include:

[0061] S101. Obtain the original image and determine the normalized pixel value of each pixel in the original image;

[0062] S102. Obtain the pixel block centered on each pixel, and determine the pixel point parameter of the pixel and the pixel block parameter of the pixel block based on the normalized pixel value;

[0063] S103. Determine the adaptive gain coefficient and adaptive correction value of the pixel based on the pixel point parameters and pixel block parameters;

[0064] S104. Determine the corrected pixel value of each pixel based on the adaptive gain coefficient and the adaptive correction value, and obtain the enhanced image based on the corrected pixel value of each pixel.

[0065] Compared with the prior art, the low-light image enhancement method provided in this embodiment of the invention determines the adaptive gain coefficient and adaptive correction value of each pixel based on pixel point parameters and pixel block parameters, and determines the corrected pixel value of each pixel based on the adaptive gain coefficient and adaptive correction value, and obtains the enhanced image based on the corrected pixel value of each pixel. This achieves the purpose of adjusting the pixel value of each pixel in the low-light image based on the adaptive gain coefficient and adaptive correction value, thereby avoiding the technical problems of amplified image noise, overexposure in high-illuminance areas, low contrast, and failure to reflect detailed features caused by uniform adjustment of each pixel in the low-light image, thus improving the image quality of the enhanced image.

[0066] Furthermore, compared to existing image enhancement methods that involve large computational demands such as deep learning, the embodiments of the present invention only require determining the adaptive gain coefficient and adaptive correction value of each pixel to obtain the enhanced image, which requires less computation and can improve the processing speed of low-light images while reducing hardware requirements.

[0067] It should be understood that the method of acquiring the original image in step S101 can be to acquire the original image in real time through an image acquisition device, or to read or call the original image from a storage medium that stores historically acquired original images.

[0068] The original image is in RGB format.

[0069] In some embodiments of the present invention, step S101 includes:

[0070] S201. Obtain the original red component pixel value, original green component pixel value, original blue component pixel value, and red component normalization coefficient, green component normalization coefficient, and blue component normalization coefficient of each pixel in the original image.

[0071] S202. Determine the normalized pixel value of each pixel based on the original red component pixel value, the original green component pixel value, the original blue component pixel value, the red component normalization coefficient, the green component normalization coefficient, and the blue component normalization coefficient.

[0072] The embodiments of the present invention determine the normalized pixel value by using the original red component pixel value, the original green component pixel value, and the original blue component pixel value. This takes into account the correlation of the R, G, and B channel colors in the original image, ensuring the color consistency between the original image and the determined enhanced image, thereby making the determined enhanced image have a more comfortable viewing experience.

[0073] Specifically, the formula for calculating the normalized pixel value of a pixel with x-coordinate and y-coordinate is as follows:

[0074] D1(x,y)=q1*R(x,y)+q2*G(x,y)+q3*B(x,y)

[0075] In the formula, D1(x,y) is the normalized pixel value; R(x,y) is the original red component pixel value; q1 is the red component normalization coefficient; G(x,y) is the original green component pixel value; q2 is the green component normalization coefficient; B(x,y) is the original blue component pixel value; and q3 is the blue component normalization coefficient.

[0076] It should be noted that in some embodiments of the present invention, the values ​​of q1, q2, and q3 should make the range of D1(x,y) [0,1].

[0077] In some other embodiments of the present invention, the value of D1(x,y) may also be in other ranges, and the values ​​of q1, q2, and q3 may be adjusted accordingly.

[0078] In some embodiments of the present invention, a pixel block of size [n, m] can be taken with each pixel as the center, where n is the number of pixels selected along the horizontal axis and m is the number of pixels selected along the vertical axis. In specific embodiments of the present invention, the values ​​of n and m range from 1 to 10.

[0079] In some embodiments of the present invention, pixel parameters include pixel transform values ​​obtained by gamma transformation of pixel values ​​and pixel transform derivative values; pixel block parameters include pixel block size, sum of normalized pixel values, pixel block mean, and pixel block mean square error.

[0080] This invention, by setting pixel parameters including pixel transformation values ​​after gamma transformation, can improve the intensity distribution of pixel values ​​in the original image based on the principle that gamma transformation is a non-linear transformation. This makes the determined enhanced image more consistent with the visual characteristics of the human eye and provides a more comfortable viewing experience for the enhanced image.

[0081] In some embodiments of the present invention, such as Figure 3 As shown, step S103 includes:

[0082] S301. Construct a segmented gain coefficient determination model based on pixel point parameters and pixel block parameters, and determine the adaptive gain coefficient of the pixel point based on the segmented gain coefficient determination model.

[0083] S302. Determine the maximum adaptive gain coefficient among the adaptive gain coefficients of each pixel.

[0084] S303. Determine the adaptive correction value of the pixel based on the adaptive gain coefficient, the maximum adaptive gain coefficient, and the preset correction value.

[0085] The embodiments of the present invention determine the adaptive gain coefficient of a pixel by using a segmented model based on the gain coefficient. Different adaptive gain processing can be applied to pixels that meet different conditions, thereby further improving the contrast of the image and preventing technical problems such as overexposure and oversaturation of the enhanced image, thus further improving the image quality of the enhanced image.

[0086] In a specific embodiment of the present invention, the gain coefficient segmentation determination model includes a first gain coefficient determination sub-model, a second gain coefficient determination sub-model, and a third gain coefficient determination model; then, as follows Figure 4 As shown, step S301 includes:

[0087] S401. When the average value of a pixel block is greater than or equal to the preset lower limit of a pixel, or the normalized pixel value is greater than the preset lower limit of a pixel and less than or equal to the preset upper limit of a pixel, the adaptive gain coefficient of the pixel is determined based on the first gain coefficient and the sub-model is determined.

[0088] S402. When the average value of the pixel block is less than the preset lower limit of the pixel value, and the normalized pixel value is greater than the average value of the pixel block and less than or equal to the preset lower limit of the pixel value, the adaptive gain coefficient of the pixel point is determined based on the second gain coefficient determination sub-model.

[0089] S403. When the average value of the pixel block is less than the preset lower limit of the pixel value, and the normalized pixel value is less than or equal to the average value of the pixel block, the adaptive gain coefficient of the pixel is determined based on the third gain coefficient determination sub-model.

[0090] Specifically, the model for determining the gain coefficient piecewise is as follows:

[0091]

[0092] In the formula, d1(x,y) is the adaptive gain coefficient of the pixel with x-coordinate and y-coordinate; D1(x,y) gamma is the pixel transformation value; size is the size of the pixel block; SUM is the sum of normalized pixel values; D1'(x,y) gamma is the derivative value of the pixel transformation; gamma is the gamma transformation coefficient; mean is the pixel block mean; MSE is the pixel block mean square error; T1 is the preset lower limit value of pixels; T2 is the preset upper limit value of pixels.

[0093] It is understandable that the first gain coefficient determines the sub-model as follows:

[0094]

[0095] The second gain coefficient determines the sub-model as follows:

[0096]

[0097] The third gain coefficient determines the sub-model as follows:

[0098] D1'(x,y) gamma / mean

[0099] In a specific embodiment of the present invention, the adaptive correction value d2(x,y) of the pixel with horizontal coordinate x and vertical coordinate y is:

[0100]

[0101] In the formula, k is a constant, with a value range of 1 to 50; ε is the gain adjustment coefficient, with a value range of -10 to 10; and max(d1) is the maximum adaptive gain coefficient.

[0102] As can be seen from the above formula for calculating the adaptive correction value, the adaptive correction value varies depending on the adaptive gain coefficient. The magnitude of the adaptive gain coefficient indicates whether the pixel it represents is a boundary pixel or not. Therefore, by setting the adaptive correction value, this embodiment of the invention can further highlight the detailed features of the enhanced image and improve the image quality of the enhanced image.

[0103] In some embodiments of the present invention, such as Figure 5 As shown, step S104 includes:

[0104] S501. Determine the red component correction pixel value of the pixel based on the original red component pixel value, the adaptive gain coefficient, and the adaptive correction value.

[0105] S502. Determine the corrected green component pixel value of a pixel based on the original green component pixel value, the adaptive gain coefficient, and the adaptive correction value.

[0106] S503. Determine the blue component correction pixel value of the pixel based on the original blue component pixel value, the adaptive gain coefficient, and the adaptive correction value.

[0107] In this embodiment of the invention, the red component correction pixel value, green component correction pixel value, and blue component correction pixel value are obtained by adaptive gain coefficient and adaptive correction value, respectively, thereby obtaining an enhanced image. This can further improve the color consistency between the enhanced image and the original image, thereby further improving the visual comfort of the enhanced image.

[0108] In a specific embodiment of the present invention, the red component correction pixel value, the green component correction pixel value, and the blue component correction pixel value are respectively:

[0109] R_bright(x,y)==d1(x,y)*R(x,y)+d2(x,y)

[0110] G_bright(x,y)==d1(x,y)*G(x,y)+d2(x,y)

[0111] B_bright(x,y)==d1(x,y)*B(x,y)+d2(x,y)

[0112] In the formula, R_bright(x,y) is the red component correction pixel value; G_bright(x,y) is the green component correction pixel value; B_bright(x,y) is the blue component correction pixel value; R(x,y) is the original red component pixel value; G(x,y) is the original green component pixel value; and B(x,y) is the original blue component pixel value.

[0113] To further improve the clarity of detailed features in the enhanced image and reduce noise, in some embodiments of the present invention, such as Figure 6 As shown, low-light image enhancement methods also include:

[0114] S601. Obtain the original pixel block centered on each pixel point in the enhanced image, and determine the set of adaptive gain coefficients corresponding to the original pixel block. Determine whether there is at least one adaptive gain coefficient in the set of adaptive gain coefficients that is greater than the gain coefficient threshold.

[0115] S602. If all adaptive gain coefficients in the adaptive gain coefficient set are less than or equal to the gain coefficient threshold, then the target pixel block is determined with each pixel in the enhanced image as the center, and the target pixel block is filtered. The center pixel of the filtered target pixel block replaces the center pixel of the original pixel block to obtain the optimized image.

[0116] S603. If at least one adaptive gain coefficient in the set of adaptive gain coefficients is greater than the gain coefficient threshold, then the target pixel block is determined with each pixel in the enhanced image as the center, and edge enhancement processing is performed on the target pixel block. The center pixel of the edge-enhanced target pixel block replaces the center pixel of the original pixel block to obtain the optimized image.

[0117] It should be understood that the adaptive gain coefficient set in step S601 includes the adaptive gain coefficients of each pixel in the original pixel block.

[0118] This invention, through obtaining the original pixel block centered on each pixel in the enhanced image and the corresponding set of adaptive gain coefficients, determines whether each pixel in the original pixel block belongs to an image detail feature pixel by judging whether at least one adaptive gain coefficient in the set is greater than a gain coefficient threshold. When all adaptive gain coefficients in the set are less than or equal to the gain coefficient threshold, the pixels in the original image block do not belong to image detail feature pixels, and filtering is applied to the target pixel block at that pixel to reduce noise. When at least one adaptive gain coefficient in the set is greater than the gain coefficient threshold, the pixels in the original image block with adaptive gain coefficients greater than the gain coefficient threshold belong to image detail feature pixels, no filtering is applied to those pixels, preserving image detail features, and edge enhancement is used to further strengthen the image detail features in the enhanced image. This further reduces noise in the obtained optimized image and further strengthens the detail features in the optimized image.

[0119] It should also be understood that the gain coefficient threshold can be set or adjusted according to the actual application scenario or empirical value. In some embodiments of the present invention, the value range of the gain coefficient threshold is 1 to max(d1), specifically, the gain coefficient threshold is the midpoint between 1 and max(d1).

[0120] It should be noted that the size of the target pixel block should also be selected based on the actual application scenario or empirical values.

[0121] It should also be noted that the filtering method in step S602 can be at least one of the following: mean filtering, Gaussian filtering, median filtering, and bilateral filtering. The edge enhancement method in step S603 can be at least one of the following: wavelet transform sharpening, Gaussian edge-preserving sharpening, etc.

[0122] To avoid the technical problem of changes in saturation and color in the enhanced and / or optimized images obtained after enhancement processing, resulting in significant differences from the original images, in some embodiments of the present invention, such as... Figure 7 As shown, low-light image enhancement methods also include:

[0123] S701. Perform saturation correction processing on the enhanced image and / or optimized image to obtain a saturation correction image;

[0124] S702. Perform pseudo-color correction processing on the saturation correction image to obtain the color correction image.

[0125] By sequentially performing saturation correction and pseudo-color correction processing on the enhanced and / or optimized images, this invention ensures that the saturation and color of the obtained color-corrected image are the same as the original image, thereby further improving the visual comfort of the color-corrected image.

[0126] In some embodiments of the present invention, step S701 specifically involves: determining the saturation value of each pixel in the enhanced image and / or optimized image, and correcting this saturation value to be the same as the saturation value of the corresponding pixel in the original image. Step S702 specifically involves: performing pseudo-color correction processing on the saturation correction image based on a preset pseudo-color correction processing method. The pseudo-color correction processing method can be any pseudo-color correction processing method in the prior art, and will not be elaborated here.

[0127] To verify the superiority of the low-light image enhancement method proposed in this embodiment of the invention, an image enhancement method based on HSV domain gamma transform is compared with the low-light image enhancement method proposed in this embodiment of the invention. Figure 8 As shown, Figure 8 The left side shows the original image, the middle side shows the enhanced image obtained after gamma transform based on the HSV domain, and the right side shows the enhanced image based on the low-light image enhancement method proposed in this application. It can be seen that the low-light image enhancement method proposed in this embodiment of the invention has a higher signal-to-noise ratio and image contrast compared with the image enhancement method based on gamma transform in the HSV domain, and generates less noise.

[0128] To better implement the low-light image enhancement method in this embodiment of the invention, based on the low-light image enhancement method, correspondingly, as follows: Figure 9As shown, this embodiment of the invention also provides a low-light image enhancement device, the low-light image enhancement device 900 comprising:

[0129] The normalized pixel value determination unit 901 is used to acquire the original image and determine the normalized pixel value of each pixel in the original image.

[0130] The parameter determination unit 902 is used to obtain pixel blocks centered on each pixel point, and determine the pixel point parameters of the pixel point and the pixel block parameters of the pixel block based on the normalized pixel values.

[0131] Gain coefficient and correction value determination unit 903 is used to determine the adaptive gain coefficient and adaptive correction value of a pixel based on pixel point parameters and pixel block parameters;

[0132] The enhanced image determination unit 904 is used to determine the corrected pixel value of a pixel based on the adaptive gain coefficient and the adaptive correction value, and to obtain an enhanced image based on the corrected pixel value of the pixel.

[0133] The low-light image enhancement device 900 provided in the above embodiments can realize the technical solutions described in the embodiments of the low-light image enhancement method. The specific implementation principles of each module or unit can be found in the corresponding content in the embodiments of the low-light image enhancement method, and will not be repeated here.

[0134] like Figure 10 As shown, the present invention also provides an image processing device 1000. The image processing device 1000 includes a processor 1001, a memory 1002, and a display 1003. Figure 10 Only some components of the image processing device 1000 are shown; however, it should be understood that it is not required to implement all of the shown components, and more or fewer components may be implemented instead.

[0135] In some embodiments, processor 1001 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 1002 or process data, such as the low-light image enhancement method of the present invention.

[0136] In some embodiments, processor 1001 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 1001 may be local or remote. In some embodiments, processor 1001 may be implemented on a cloud platform. In some embodiments of the present invention, the cloud platform may include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, inter-cloud, multi-cloud, etc., or any combination thereof.

[0137] In some embodiments, memory 1002 may be an internal storage unit of the image processing device 1000, such as a hard disk or memory of the image processing device 1000. In other embodiments, memory 1002 may also be an external storage device of the image processing device 1000, such as a pluggable hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., provided on the image processing device 1000.

[0138] Furthermore, the memory 1002 may include both internal storage units of the image processing device 1000 and external storage devices. The memory 1002 is used to store application software and various types of data for which the image processing device 1000 is installed.

[0139] In some embodiments, display 1003 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 1003 is used to display information from the image processing device 1000 and to display a visual user interface. Components 1001-1003 of the image processing device 1000 communicate with each other via a system bus.

[0140] In some embodiments of the present invention, when the processor 1001 executes the low-light image enhancement program in the memory 1002, the following steps may be performed:

[0141] Acquire the original image and determine the normalized pixel value of each pixel in the original image;

[0142] Obtain pixel blocks centered on each pixel, and determine the pixel parameters of the pixel points and the pixel block parameters of the pixel blocks based on the normalized pixel values;

[0143] The adaptive gain coefficient and adaptive correction value of the pixel are determined based on the pixel point parameters and pixel block parameters;

[0144] The corrected pixel value of each pixel is determined based on the adaptive gain coefficient and the adaptive correction value, and the enhanced image is obtained based on the corrected pixel value of each pixel.

[0145] Furthermore, this embodiment of the invention does not specifically limit the type of the image processing device 1000 mentioned. The image processing device 1000 can be a portable image processing device such as a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, or laptop computer. Exemplary embodiments of portable image processing devices include, but are not limited to, portable image processing devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable image processing device can also be other portable image processing devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the invention, the image processing device 1000 may not be a portable image processing device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0146] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0147] The low-light image enhancement method, apparatus, and image processing device provided by the present invention have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A low-light image enhancement method, characterized in that, include: Acquire the original image and determine the normalized pixel value of each pixel in the original image; Obtain pixel blocks centered on each of the aforementioned pixels, and determine the pixel point parameters of the pixels and the pixel block parameters of the pixel blocks based on the normalized pixel values; A gain coefficient segmentation determination model is constructed based on the pixel point parameters and the pixel block parameters, and the adaptive gain coefficient of the pixel point is determined based on the gain coefficient segmentation determination model. Determine the maximum adaptive gain coefficient among the adaptive gain coefficients of each pixel; The adaptive correction value of the pixel is determined by the model based on the adaptive gain coefficient, the maximum adaptive gain coefficient, and the preset correction value. The corrected pixel value of the pixel is determined based on the adaptive gain coefficient and the adaptive correction value, and the enhanced image is obtained based on the corrected pixel value of the pixel.

2. The low-light image enhancement method according to claim 1, characterized in that, Determining the normalized pixel value of each pixel in the original image includes: Obtain the original red component pixel value, original green component pixel value, original blue component pixel value, and red component normalization coefficient, green component normalization coefficient, and blue component normalization coefficient of each pixel in the original image; The normalized pixel value of each pixel is determined based on the original red component pixel value, the original green component pixel value, the original blue component pixel value, the red component normalization coefficient, the green component normalization coefficient, and the blue component normalization coefficient.

3. The low-light image enhancement method according to claim 1, characterized in that, The pixel parameters include the pixel transform value after gamma transformation of the pixel value and the pixel transform derivative value; the pixel block parameters include the size of the pixel block, the sum of normalized pixel values, the pixel block mean, and the pixel block mean square error.

4. The low-light image enhancement method according to claim 3, characterized in that, The gain coefficient segmentation determination model includes a first gain coefficient determination sub-model, a second gain coefficient determination sub-model, and a third gain coefficient determination sub-model; The step of determining the adaptive gain coefficient of the pixel based on the gain coefficient segmentation determination model includes: When the mean value of the pixel block is greater than or equal to the preset lower limit value, or the normalized pixel value is greater than the preset lower limit value and less than or equal to the preset upper limit value, the adaptive gain coefficient of the pixel is determined based on the first gain coefficient determination sub-model. When the average value of the pixel block is less than the preset lower limit value, and the normalized pixel value is greater than the average value of the pixel block but less than or equal to the preset lower limit value, the adaptive gain coefficient of the pixel is determined based on the second gain coefficient determination sub-model. When the average value of the pixel block is less than the preset lower limit value, and the normalized pixel value is less than or equal to the average value of the pixel block, the adaptive gain coefficient of the pixel is determined based on the third gain coefficient determination sub-model.

5. The low-light image enhancement method according to claim 2, characterized in that, Determining the corrected pixel value of the pixel based on the adaptive gain coefficient and the adaptive correction value includes: The red component correction pixel value of the pixel is determined based on the original red component pixel value, the adaptive gain coefficient, and the adaptive correction value; The green component corrected pixel value of the pixel is determined based on the original green component pixel value, the adaptive gain coefficient, and the adaptive correction value; The blue component correction pixel value of the pixel is determined based on the original blue component pixel value, the adaptive gain coefficient, and the adaptive correction value.

6. The low-light image enhancement method according to claim 1, characterized in that, The low-light image enhancement method further includes: Obtain the original pixel block centered on each pixel in the enhanced image, determine the set of adaptive gain coefficients corresponding to the original pixel block, and determine whether there is at least one adaptive gain coefficient in the set of adaptive gain coefficients that is greater than the gain coefficient threshold. If all adaptive gain coefficients in the set of adaptive gain coefficients are less than or equal to the gain coefficient threshold, then a target pixel block is determined with each pixel in the enhanced image as the center, and the target pixel block is filtered. The center pixel of the filtered target pixel block replaces the center pixel of the original pixel block to obtain an optimized image. If at least one adaptive gain coefficient in the set of adaptive gain coefficients is greater than the gain coefficient threshold, then a target pixel block is determined with each pixel in the enhanced image as the center, and edge enhancement processing is performed on the target pixel block. The center pixel of the edge-enhanced target pixel block replaces the center pixel of the original pixel block to obtain an optimized image.

7. The low-light image enhancement method according to claim 6, characterized in that, The low-light image enhancement method further includes: The enhanced image and / or the optimized image are subjected to saturation correction processing to obtain a saturation correction image; The saturation correction image is subjected to pseudo-color correction processing to obtain a color correction image.

8. A low-light image enhancement device, characterized in that, include: A normalized pixel value determination unit is used to acquire the original image and determine the normalized pixel value of each pixel in the original image. A parameter determination unit is used to obtain a pixel block centered on each pixel point, and to determine the pixel point parameter of the pixel point and the pixel block parameter of the pixel block based on the normalized pixel value; The gain coefficient and correction value determination unit is used to construct a gain coefficient segmentation determination model based on the pixel point parameters and the pixel block parameters, and determine the adaptive gain coefficient of the pixel point based on the gain coefficient segmentation determination model; and determine the maximum adaptive gain coefficient among the adaptive gain coefficients of each pixel point. The adaptive correction value of the pixel is determined by the model based on the adaptive gain coefficient, the maximum adaptive gain coefficient, and the preset correction value. An enhanced image determination unit is used to determine the corrected pixel value of the pixel based on the adaptive gain coefficient and the adaptive correction value, and to obtain an enhanced image based on the corrected pixel value of the pixel.

9. An image processing device, characterized in that, Including memory and processor, among which, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the low-light image enhancement method according to any one of claims 1 to 7.

Images