CN116152094B - Automatic driving high dynamic scene image optimization method, system, terminal and medium - Google Patents
Automatic driving high dynamic scene image optimization method, system, terminal and medium Download PDFInfo
- Publication number
- CN116152094B CN116152094B CN202310024528.1A CN202310024528A CN116152094B CN 116152094 B CN116152094 B CN 116152094B CN 202310024528 A CN202310024528 A CN 202310024528A CN 116152094 B CN116152094 B CN 116152094B
- Authority
- CN
- China
- Prior art keywords
- image
- value
- brightness
- pixel
- mapping
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 88
- 238000005457 optimization Methods 0.000 title claims abstract description 27
- 238000013507 mapping Methods 0.000 claims abstract description 156
- 230000008569 process Effects 0.000 claims abstract description 22
- 230000006872 improvement Effects 0.000 claims abstract description 9
- 241000023320 Luma <angiosperm> Species 0.000 claims description 22
- 230000015654 memory Effects 0.000 claims description 22
- OSWPMRLSEDHDFF-UHFFFAOYSA-N methyl salicylate Chemical compound COC(=O)C1=CC=CC=C1O OSWPMRLSEDHDFF-UHFFFAOYSA-N 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000001914 filtration Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 11
- 239000006185 dispersion Substances 0.000 claims description 11
- 230000002093 peripheral effect Effects 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 6
- 230000002146 bilateral effect Effects 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 238000000540 analysis of variance Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 9
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Classifications
-
- G06T5/92—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration by the use of histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20208—High dynamic range [HDR] image processing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Facsimile Image Signal Circuits (AREA)
Abstract
The invention provides a method and a system for optimizing an automatic driving high-dynamic scene image, wherein the method comprises the following steps: acquiring a brightness value of each pixel point in an automatic driving scene image; counting the brightness value of each pixel point in the image to obtain histogram data; judging the class of the automatic driving scene according to the histogram data and obtaining coordinates of the regulation points of the mapping curve to obtain the mapping curve; and according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, carrying out pixel-by-pixel mapping processing on the image, realizing the dynamic range improvement on the automatic driving high dynamic scene image, and completing the optimization on the automatic driving high dynamic scene image. The invention can improve the brightness of the low-brightness area of the image and ensure that the high-brightness area is not affected by other factors; the problem that noise is excessively amplified in the mapping process of the image can be effectively avoided; the problem of saturation anomaly can be avoided during image mapping.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an automatic driving high-dynamic scene image optimization method, an automatic driving high-dynamic scene image optimization system, a terminal and a medium.
Background
The core of automatic driving is that the human driving automobile is replaced by the fusion of a computer and an artificial intelligence technology, so that the automobile can automatically complete, safe and effective driving behaviors. The safe and reliable perception capability of the automatic driving automobile in a social application scene is indispensable, so that not only is stronger calculation force needed, but also more abundant environmental information is needed to be obtained through the imaging processing technology with high resolution and high frame rate. Currently, because of limitations of the bit width of the display device or the image processing power, an HDR (high dynamic range) image needs to be displayed on an LDR (low dynamic range) device, and if the HDR image is directly shifted, a great deal of information is lost, so that more dark and bright details cannot be displayed.
In order to improve the dynamic range of the image mapping process, the prior art usually directly and greatly improves all brightness intervals of the whole image, and although the brightness of a low-brightness area can be improved, the high-brightness area can also cause overexposure and problem of losing due to the great improvement of the brightness. In the prior art, the original image is directly brought into the mapping curve to obtain the corresponding mapping image, and the influence caused by noise is not considered, so that the noise is excessively amplified. In addition, in the mapping process, the original value of the pixel point of the image is usually directly brought into the mapping curve to obtain the final mapping output value in the prior art, but peripheral pixel points are not considered, the proportion between the corresponding pixel channels of the mapped pixels is actually changed, and the serious saturation abnormality problem of the whole image is easily caused.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides an automatic driving high dynamic range image optimization method, an automatic driving high dynamic range image optimization system, a terminal and a medium.
According to one aspect of the present invention, there is provided an autopilot high dynamic scene image optimization method comprising:
acquiring a brightness value of each pixel point in an automatic driving scene image;
counting the brightness value of each pixel point in the image to obtain histogram data; the histogram data comprises a brightness histogram and the number of pixels corresponding to a brightness interval;
judging the category of the automatic driving scene according to the histogram data, and obtaining coordinates of the regulation points of the mapping curve to obtain the mapping curve;
and according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, carrying out pixel-by-pixel mapping processing on the image, realizing the dynamic range improvement on the automatic driving high dynamic scene image, and completing the optimization on the automatic driving high dynamic scene image.
Optionally, the acquiring the brightness value of each pixel point in the autopilot scene image includes:
according to the format type of the automatic driving scene image, a corresponding brightness value calculation method is adopted to obtain the brightness value of each pixel point in the automatic driving scene image; wherein:
The format type of the automatic driving scene image comprises: RAW, YUV, and RGB.
Optionally, for the automatic driving scene image format of the RAW, the brightness value calculating method includes any one or more of the following:
-performing pixel-by-pixel original image statistics on the automatic driving scene image by using a block, wherein the maximum value, the average value or the average value of the maximum value and the minimum value of the pixel values in the block with the set pixel size taking the current pixel as the center is used as the brightness value of the current pixel;
-performing difference processing on the block of the set pixel size corresponding to each pixel point by using an interpolation method to obtain pseudo R ', G ' and B ' color values of the current pixel point; and calculating the color values of the pseudo R ', G' and B ', and taking the maximum value, the average value of the maximum value and the minimum value or the weighted value of the color values of the pseudo R', G 'and B' as the brightness value of the current pixel point.
Optionally, for YUV, the luminance value calculation method includes:
and obtaining Y channel data corresponding to each pixel point, and calculating by adopting a block with a set pixel size taking the current pixel point as the center to obtain a corresponding brightness value.
Optionally, for the RGB autopilot scene image format, the luminance value calculation method includes:
obtaining the brightness value of the current pixel point by adopting a weighting method;
and weighting the brightness value of the current pixel point and the brightness values of the peripheral pixel points by using a block with the set pixel size to obtain the final brightness value of the current pixel point.
Optionally, the counting the brightness value of each pixel in the image to obtain histogram data includes:
dividing the brightness range of the image into a plurality of brightness intervals;
carrying out brightness histogram statistics according to the brightness value of each pixel in the brightness interval;
and obtaining the number of pixels corresponding to each brightness interval according to the result of the brightness histogram statistics.
Optionally, the determining the class of the automatic driving scene according to the histogram data and obtaining coordinates of the regulation points of the mapping curve to obtain the mapping curve includes:
obtaining a low-brightness value, an average value and a high-brightness value of a current image according to the histogram data, wherein the low-brightness value represents a brightness value corresponding to a low-brightness pixel of the image, the high-brightness value represents a brightness value corresponding to a high-brightness pixel of the image, and the average value is the average value of brightness sums of all pixels in the image;
Analyzing a degree of dispersion of the luminance histogram, the degree of dispersion characterizing a dynamic range of the image;
counting the proportion of the number of pixels corresponding to the brightness interval to the total number of the image pixels;
judging whether the current image is a high dynamic scene image or not according to the ambient brightness value of the image, the difference value between the low brightness value and the high brightness value, the discrete degree of the brightness histogram and the proportion of the number of pixels corresponding to the brightness interval to the total number of pixels of the image; if yes, the control intensity of the mapping curve control points corresponding to the image is given, and a mapping curve is obtained; if not, dynamic range boosting is not required.
Optionally, analyzing the degree of dispersion of the luminance histogram in the histogram data, using an analysis of variance method, includes:
wherein s is 2 Is variance, used to represent the degree of dispersion of luminance histogram, t is the number of luminance intervals of the input image, x [ i ]]For the luminance value corresponding to the ith luminance segment, hist [ i ]]And n is the total number of pixels of the image, wherein the number corresponds to the ith brightness interval.
Optionally, the method for judging whether the current image is a high dynamic scene image includes:
And when the environment brightness value is more than or equal to a set environment brightness threshold value, the difference value is more than or equal to a set difference threshold value, the discrete degree is more than or equal to a set discrete threshold value and the proportion is more than or equal to a set proportion threshold value, judging that the image is a high dynamic scene image.
Optionally, the mapping process of pixel-by-pixel is performed on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, including:
separating the high-frequency information in the brightness histogram to obtain low-frequency information in the brightness histogram;
based on the low-frequency information, mapping each pixel point of the image point by point according to a low brightness value, a high brightness value and a mapping curve of the whole image to obtain a mapping output value;
and adding the separated high-frequency information back to the mapping output value to realize the dynamic range improvement of the automatic driving high-dynamic scene image.
Optionally, the frequency division method for separating high-frequency information in the luminance histogram includes: gaussian filtering, weighted least squares filtering, guided filtering, bilateral filtering and wavelet transformation.
Optionally, based on the low-frequency information, according to a low-brightness value, a high-brightness value and a mapping curve of the whole image, mapping each pixel point of the image point by point to obtain a mapping output value, including:
Y out ＝f(luma_base) (10)
wherein, luma_base is the image low-frequency brightness value, f (luma_base) is the function corresponding to the image low-frequency brightness mapping curve, Y out An output value obtained according to the mapping curve for the low-frequency information lumabase;
wherein MaxOut is the maximum value of brightness range after image bit width reduction, Y min Is of low brightness value, Y max Is a highlight value, Y out The temp is an output value corresponding to the mapping of the low-frequency information lumabase to the output bit width;
wherein pix _ in is the picture original data channel value,and as the brightness gain value, pix_out is the output value after the bit width reduction corresponding to the channel value, namely the mapping output value.
Optionally, the adding the separated high frequency information back to the mapping output value includes:
calculating a high-frequency back-adding coefficient;
and according to the high-frequency back-adding coefficient, carrying out back-adding on the separated high-frequency information to the mapping output value pix_out to obtain a final mapping output value pixel_out, wherein the final mapping output value pixel_out is as follows:
Pixel_out＝pix_out+lumadetail*r(lumadetail，luma) (13)
where r (luma) is a high-frequency back-adding coefficient, luma is high-frequency information, and luma is a pixel brightness value.
Optionally, the high-frequency back-adding coefficient is obtained by any one of the following modes:
-setting a fixed constant;
-taking the absolute value of the difference between the high frequency information and the pixel brightness value as a high frequency back-addition coefficient function factor, and obtaining the corresponding coefficient value through a back-addition coefficient lookup table according to the set difference value threshold.
According to another aspect of the present invention, there is provided an autopilot high dynamic scene image optimization system comprising:
the pixel brightness value acquisition module is used for acquiring the brightness value of each pixel in the automatic driving scene image;
the histogram data acquisition module is used for counting the brightness value of each pixel point in the image to obtain histogram data, and the histogram data comprises a brightness histogram and the number of pixels corresponding to a brightness interval;
the mapping curve acquisition module is used for judging the category of the automatic driving scene according to the histogram data and acquiring coordinates of the regulation points of the mapping curve to obtain the mapping curve;
and the mapping processing module is used for carrying out pixel-by-pixel mapping processing on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, so as to realize the dynamic range improvement of the automatic driving high dynamic scene image and finish the optimization of the automatic driving high dynamic scene image.
According to a third aspect of the present invention there is provided a computer terminal comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program being operable to perform the method of any one of the preceding claims or to run the system of the preceding claims.
According to a fourth aspect of the present invention there is provided a computer readable storage medium having stored thereon a computer program which when executed by a processor is operable to perform a method of any of the above, or to run a system as described above.
Due to the adoption of the technical scheme, compared with the prior art, the invention has at least one of the following beneficial effects:
according to the method, the system, the terminal and the medium for optimizing the image of the automatic driving high dynamic range, the image mapping curve acquisition method adopts the mapping curve controlled by multiple points and the self-adaptive dynamic range lifting mode, so that the regulation and control are flexible, the brightness of a low-brightness area of the image can be improved, the high-brightness area is not affected by other influences, and the mapping method is flexible and has good self-adaptation.
The method, the system, the terminal and the medium for optimizing the automatic driving high dynamic range image, which are provided by the invention, can effectively avoid the problem that the noise is excessively amplified in the image mapping process.
According to the method, the system, the terminal and the medium for optimizing the automatic driving high dynamic range image, the brightness value of the current pixel is kept small in difference with the gain of the peripheral pixel in the mapping process, the saturation is basically not influenced, and the problem of abnormal saturation in the image mapping process can be avoided.
The method, the system, the terminal and the medium for optimizing the automatic driving high dynamic range image, provided by the invention, consider the parameter values in multiple aspects, not only comprise the average brightness information value, the maximum brightness information value and the minimum brightness information value of the scene, but also comprise the discrete degree of the brightness histogram and the pixel number occupation ratio of each brightness interval, can more particularly classify and judge the image scene category, is applicable to different mapping curves for different scene categories, and is flexible and changeable.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the detailed description of non-limiting embodiments, given with reference to the accompanying drawings in which:
fig. 1 is a workflow diagram of an automatic driving high dynamic scene image optimization method according to an embodiment of the present invention.
Fig. 2 is a schematic working diagram of an automatic driving high dynamic scene image optimization method according to a preferred embodiment of the invention.
FIG. 3 is a block diagram of a pixel size of 5×5 in a preferred embodiment of the present invention.
FIG. 4 is a block diagram of Y channel 3X3 in a preferred embodiment of the present invention.
FIG. 5 is a schematic diagram of a mapping curve composed of multiple linear curves according to a preferred embodiment of the present invention.
FIG. 6 is a schematic diagram of a mapping curve composed of linear and nonlinear curves in accordance with a preferred embodiment of the present invention.
FIG. 7 is a schematic diagram of the high frequency back-addition coefficients in a preferred embodiment of the present invention.
Fig. 8 is a schematic diagram of a composition module of an automatic driving high dynamic scene image optimization system according to an embodiment of the invention.
Detailed Description
The following describes embodiments of the present invention in detail: the embodiment is implemented on the premise of the technical scheme of the invention, and detailed implementation modes and specific operation processes are given. It should be noted that variations and modifications can be made by those skilled in the art without departing from the spirit of the invention, which falls within the scope of the invention.
The embodiment of the invention provides an automatic driving high-dynamic scene image optimization method, which adopts a multi-point control mapping curve and a self-adaptive dynamic range lifting mode by two large processes of mapping curve acquisition and mapping treatment, so that the brightness of a low-brightness area of an image can be lifted, the high-brightness area is not influenced by other factors, and the automatic driving high-dynamic scene image is optimized.
As shown in fig. 1, the method for optimizing the autopilot high dynamic scene image provided in this embodiment may include:
s1, acquiring a brightness value of each pixel point in an automatic driving scene image;
s2, counting the brightness value of each pixel point in the image to obtain histogram data; the histogram data comprises a brightness histogram and the number of pixels corresponding to a brightness interval;
s3, judging the category of the automatic driving scene according to the histogram data, and obtaining coordinates of the regulation points of the mapping curve to obtain the mapping curve;
and S4, carrying out pixel-by-pixel mapping processing on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, so as to realize the dynamic range improvement of the automatic driving high dynamic scene image and complete the optimization of the automatic driving high dynamic scene image.
In a preferred embodiment of S1, acquiring the brightness value of each pixel in the autopilot scene image may include:
according to the format type of the automatic driving scene image, a corresponding brightness value calculation method is adopted to obtain the brightness value of each pixel point in the automatic driving scene image; wherein:
a format type of an autopilot scene image, comprising: RAW, YUV, and RGB.
In a preferred embodiment of S1, the luminance value calculation method may include any one or more of the following for the automatic driving scene image format of RAW:
-performing pixel-by-pixel original image statistics on the automatic driving scene image by using a block, wherein the maximum value, the average value or the average value of the maximum value and the minimum value of the pixel values in the block with the set pixel size taking the current pixel as the center is used as the brightness value of the current pixel;
-performing difference processing on the block of the set pixel size corresponding to each pixel point by using an interpolation method to obtain pseudo R ', G ' and B ' color values of the current pixel point; and calculating the color values of the pseudo R ', G' and B ', and taking the maximum value, the average value of the maximum value and the minimum value or the weighted value of the color values of the pseudo R', G 'and B' as the brightness value of the current pixel point.
In a preferred embodiment of S1, for YUV autopilot scene image format, the luminance value calculation method may include:
and obtaining Y channel data corresponding to each pixel point, and calculating by adopting a block with a set pixel size taking the current pixel point as the center to obtain a corresponding brightness value.
In a preferred embodiment of S1, the luminance value calculation method may include, for the RGB autopilot image format:
Obtaining the brightness value of the current pixel point by adopting a weighting method;
and weighting the brightness value of the current pixel point and the brightness values of the peripheral pixel points by using a block with the set pixel size to obtain the final brightness value of the current pixel point.
In a preferred embodiment of S2, the statistics of the brightness value of each pixel in the image to obtain histogram data may include:
s21, dividing the brightness range of the image into a plurality of brightness intervals;
s22, carrying out brightness histogram statistics according to the brightness value of each pixel in the brightness interval;
s23, obtaining the number of pixels corresponding to each brightness interval according to the result of the brightness histogram statistics.
In S21, the luminance range is determined according to the bit width of the image, for example, 8bit data, and the luminance range of the image is 0-255.
In a preferred embodiment of S3, determining the class of the automatic driving scene according to the histogram data and obtaining coordinates of the control points of the mapping curve to obtain the mapping curve may include:
s31, obtaining a low-brightness value, an average value and a high-brightness value of a current image according to histogram data, wherein the low-brightness value represents a brightness value corresponding to a low-brightness pixel of the image, the high-brightness value represents a brightness value corresponding to a high-brightness pixel of the image, the average value is the average value of brightness sums of all pixels in the image, and the average value provides a basis for calculation of a mapping curve;
S32, analyzing the discrete degree of the brightness histogram, wherein the discrete degree characterizes the dynamic range of the image;
s33, counting the proportion of the number of pixels corresponding to the brightness interval to the total number of pixels of the image;
s34, judging whether the current image is a high dynamic scene image according to the ambient brightness value, the difference value between the low brightness value and the high brightness value of the image, the discrete degree of the brightness histogram and the proportion of the number of pixels corresponding to the brightness interval to the total number of pixels of the image; if yes, the control intensity of the mapping curve control points corresponding to the image is given, and a mapping curve is obtained; if not, dynamic range boosting is not required.
In S34, the ambient brightness value is the ambient brightness around the camera when shooting, and is obtained by AE.
In a preferred embodiment of S32, analyzing the degree of dispersion of the luminance histogram in the histogram data, by using a variance analysis method, may include:
wherein s is 2 Is variance, used to represent the degree of dispersion of luminance histogram, t is the number of luminance intervals of the input image, x [ i ]]For the luminance value corresponding to the ith luminance segment, hist [ i ]]And n is the total number of pixels of the image, wherein the number corresponds to the ith brightness interval.
In a preferred embodiment of S34, the method for determining whether the current image is a high dynamic scene image may include:
And when the environment brightness value is more than or equal to the set environment brightness threshold value, the difference value is more than or equal to the set difference threshold value, the discrete degree is more than or equal to the set discrete threshold value and the proportion is more than or equal to the set proportion threshold value, judging that the image is a high dynamic scene image.
In S34, the control intensity is obtained by integrating the variance of the histogram data and the ambient brightness information. Different environment brightness can be configured with different regulation and control parameters, and different control intensities can be configured under the same environment brightness and different dynamic scenes, for example, under a high dynamic scene, the intensity configuration of low brightness improvement is higher.
In a preferred embodiment of S4, the mapping process of the image pixel by pixel according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image may include:
s41, separating high-frequency information in the brightness histogram to obtain low-frequency information in the brightness histogram;
s42, based on the low-frequency information, mapping each pixel point of the image point by point according to the low brightness value, the high brightness value and the mapping curve of the whole image to obtain a mapping output value;
s43, the separated high-frequency information is added back to the mapping output value, and the dynamic range of the automatic driving high-dynamic scene image is improved.
In S4, the high brightness value (high brightness value) and the low brightness value (low brightness value) are obtained by statistical analysis of the brightness histogram, and the proportion of the high brightness pixels is set to 10% according to a certain pixel number proportion, for example, 1000 pixels of the whole image, then the brightness histogram is searched from the maximum value, and the brightness value corresponding to the 100 th pixel is always searched to be the high brightness value; the same principle, a low light value is obtained. Wherein the high bright pixel and the low bright pixel are judged by a set threshold value.
In S42, the low-frequency information is a luminance value obtained by filtering the image luminance map.
In S43, the high-frequency information, that is, the luminance value of the pixel is filtered to obtain a low-frequency luminance value, and the difference between the luminance value of the pixel and the low-frequency luminance value becomes the high-frequency information value.
In a preferred embodiment of S41, the frequency division method for separating the high frequency information in the luminance histogram includes: gaussian filtering, weighted least squares filtering, guided filtering, bilateral filtering and wavelet transformation.
In a preferred embodiment of S42, based on the low frequency information, according to the low brightness value, the high brightness value and the mapping curve of the whole image, mapping each pixel point of the image point by point to obtain a mapping output value may include:
Y out ＝f(luma_base) (15)
Wherein, luma_base is the image low-frequency brightness value, f (luma_base) is the function corresponding to the image low-frequency brightness mapping curve, Y out An output value obtained according to the mapping curve for the low-frequency information lumabase;
wherein M isaxOut is the maximum value of brightness range after image bit width reduction, Y min Is of low brightness value, Y max Is a highlight value, Y out The temp is an output value corresponding to the mapping of the low-frequency information lumabase to the output bit width;
wherein pix _ in is the picture original data channel value,and as the brightness gain value, pix_out is the output value after the bit width reduction corresponding to the channel value, namely the mapping output value.
In S42, if the original image is 10 bits, the luminance range is 0-1023, and the luminance range is 0-255 after the bit width is reduced to 8 bits, the MaxOut is 255.
In a preferred embodiment of S43, adding the separated high frequency information back to the mapped output value may include:
s431, calculating a high-frequency back-adding coefficient;
s432, according to the high-frequency back-adding coefficient, the separated high-frequency information is back-added to the mapping output value pix_out, and the final mapping output value pixel_out is obtained as follows:
Pixel_out＝pix_out+lumadetail*r(lumadetail，luma) (18)
where r (luma) is a high-frequency back-adding coefficient, luma is high-frequency information, and luma is a pixel brightness value.
In a preferred embodiment of S431, the high frequency back-addition coefficient may be obtained in any of the following ways:
-setting a fixed constant;
-taking the absolute value of the difference between the high frequency information and the pixel brightness value as a high frequency back-addition coefficient function factor, and obtaining the corresponding coefficient value through a back-addition coefficient lookup table according to the set difference value threshold.
The technical scheme provided by the embodiment of the invention is further described below with reference to the accompanying drawings.
The optimization method provided by the embodiment of the invention can be divided into two steps of mapping curve acquisition and mapping process. The specific operation is shown in fig. 2.
Step 1: mapping curve acquisition
The mapping curve is obtained by analyzing the image brightness histogram data. Firstly, obtaining a brightness value of each pixel point in an image through a certain calculation mode; then, counting the brightness value of each pixel in the whole image to obtain final histogram data; finally, judging the scene category according to the histogram data and obtaining the coordinates of the regulation points of the mapping curve, thus obtaining the final mapping curve. The input image may be RAW, YUV, RGB or other data types, and the following is a detailed description of the mapping curve acquisition process.
Step 1.1: acquiring brightness values of image pixels
(1) Assume that the input image is a RAW image
The method comprises the following steps:
and analyzing the original data of the RAW image pixel by using blocks with other sizes of 3x3 or 5x5 and the like, wherein the maximum value, the average value or the average value of the maximum value and the minimum value in the block with the current pixel as the center is used as the brightness value of the current pixel. Taking the maximum value in the 3x3block as an example of the luminance value, the corresponding formula is as follows:
Luma＝MAX(block 00 ，block 01 ，......，block 22 ) (19)
the second method is as follows:
and processing blocks of other sizes such as 3x3 or 5x5 and the like corresponding to each pixel point by using an interpolation mode to obtain pseudo R ', G' and B 'values of the current pixel point, and analyzing the pseudo R', G 'and B' values of the current pixel point. As shown in FIG. 3, at B 22 Taking a 5x5block as the center as an example, the pseudo R ', G ' and B ' values of the current pixel point are obtained by interpolation.
The calculation formula of the pseudo B' value is as follows:
B`＝(B 00 +2*B 02 +B 04 +2*B 20 +4*B 22 +2*B 24 +B 40 +2*B 42 +B 44 )/16 (20)
the calculation formula of the pseudo G' value is as follows:
G`＝(G 01 +G 03 +G 10 +2*G 12 +G 14 +2*G 21 +2*G 23 +G 30 +2*G 32 +G 34 +G 41 +G 43 )/16 (21)
the calculation formula of the pseudo R' value is as follows:
R`＝(R 11 +R 13 +R 31 +R 33 )/4 (22)
the brightness value can be calculated according to the pseudo R ', G' and B 'values of the current pixel point, and the brightness value can be the maximum value of the R', G 'and B' values:
Luma＝MAX(R`，G`，B`) (23)
the luminance value may be an average of the values of R ', G ', B ':
Luma＝(R`+G`+B`)/3 (24)
the brightness value can be the average value of the maximum value and the minimum value of the R ', G ' and B ':
Luma＝(MAX(R`，G`，B`)+MIN(R`，G`，B`))/2 (25)
the luminance value may be a weighted value of the R ', G ', B ' values:
Luma＝0.299*R`+0.578*G+0.114*B (26)
(2) Assume that the input image is a YUV image
And analyzing the YUV graph to obtain Y channel data corresponding to each pixel, and calculating to obtain a corresponding brightness value by adopting a 3x3 or 5x5block taking the current pixel point as the center. As shown in fig. 4, a schematic diagram of Y channel 3x3block centered on Y11 is shown.
If the 3x3block Y channel data is weighted to a certain degree, the brightness value corresponding to the current pixel point is obtained, and the calculation formula is as follows:
Luma＝(Y 00 +2*Y 01 +Y 02 +2*Y 10 +4*Y 11 +2*Y 12 +Y 20 +2*Y 21 +Y 22 )/16 (27)
(3) assume that the input image is an RGB map
The brightness value of the current pixel can be obtained by adopting a weighting mode aiming at the RGB image, and the calculation formula is as follows:
Luma＝0.299*R+0.578*G+0.114*B (28)
the formula (28) can basically represent the brightness of the current pixel point, and the final brightness value of the current pixel point is obtained by weighting the current pixel brightness value and the peripheral pixel brightness values by using a 3x3 or 5x5block as formula (27).
Step 1.2: luminance histogram statistics
Firstly, dividing an image brightness range into a plurality of brightness intervals, wherein the number of the brightness intervals can be distributed according to actual requirements; then, carrying out brightness histogram statistics on each pixel; and finally obtaining the number of pixels corresponding to each brightness interval.
Step 1.3: analyzing the brightness histogram data to obtain a mapping curve
First, a low luminance value Y of a current image is obtained from a luminance histogram min Average value Ya vg And a highlight value Y max Wherein the brightness value Y is low min Brightness value, high brightness value Y, corresponding to low brightness pixel representing whole image max The luminance value corresponding to the highlighted pixel of the whole image is characterized. Low brightness value Y min And a highlight value Y max The value is an important factor for the whole image dynamic range mapping.
Then, the brightness histogram corresponding to the whole image is analyzed, and the discrete degree can measure the data distribution condition of one image histogram, so that the dynamic range of the whole image can be represented. There are various ways of analyzing the degree of dispersion of the luminance histogram, such as the polar difference, the average difference, the variance, the standard deviation, and the like. For example, the variance characterizes the degree of dispersion of the luminance histogram as:
wherein s is 2 Is variance, t is the number of brightness intervals divided by brightness range of input image, x [ i ]]For the luminance value corresponding to the ith interval, hist [ i ]]The number of pixels in the ith section, n is the total imageNumber of elements. If the variance s 2 The smaller the value, the more pixels are clustered in Y avg Near brightness, the image can be estimated to be a non-high dynamic scene; if the variance s 2 The larger the value, the larger the difference between most pixels and average luminance, the more the image can be estimated to be a high dynamic scene.
And secondly, counting the proportion of the number of pixels in each brightness interval. And setting some brightness thresholds to divide the whole image brightness range into a plurality of brightness intervals, and analyzing the proportion of the number of pixels in each brightness interval to the total number of pixels in the whole image. The proportion of the number of pixels in each section is also an important factor for judging whether the scene is a high dynamic scene.
Then, the scene category is analyzed. According to the ambient brightness value and low brightness value Y of the scene min And a highlight value Y max The difference value of the histogram, the discrete degree of the histogram and the proportion of the number of pixels in each brightness interval are comprehensively analyzed to judge whether the current image is a high dynamic scene image. For example, the current ambient brightness is the outdoor environment, Y min And Y is equal to max The current scene can be considered as a high dynamic scene if the value difference is large, the discrete degree of the histogram is high, and the proportion of the number of pixels in the low-brightness area is large. And if the current image is a high dynamic scene, assigning the corresponding control intensity of the control points of the mapping curve. If the scene is a low dynamic scene, the dynamic range is not required to be lifted.
Finally, a mapping curve is obtained. The mapping curve of the scheme can improve the brightness of the low-brightness area and keep the details of the high-brightness area unchanged. The mapping curve applicable to the method can be various, such as a plurality of linear straight lines, nonlinear curves and the like. For example, as shown in fig. 5, the mapping curve is composed of a plurality of linear lines, and the coordinates of the control points can be adjusted according to the dynamic range of the scene and the scene brightness, so that the brightness of the low-brightness area of the image can be improved, and the effect that the high-brightness area is not affected can be achieved.
As shown in fig. 6, the map curve is composed of a linear straight line and a nonlinear curve, which uses linear brightness increase in the very low brightness region and then uses nonlinear control in other positions. The mapping rule of the curve is controlled by four points P0 to P3, so that the brightness of a low-brightness area can be improved, and a high-brightness area is not affected by other factors.
Step 2: mapping procedure
After the step 1, the low brightness value Y of the image can be obtained min High value Y max The brightness histogram and the mapping curve, and then the whole image is required to be mapped pixel by pixel. The following is a detailed description of the mapping process.
Step 2.1 luminance map frequency division processing
In order to avoid the excessive amplification of noise in the high-frequency information in the mapping process, the high-frequency information in the image brightness map needs to be separated, only the low-frequency information is mapped, and finally the high-frequency information is added back according to a certain proportion. There are various methods for dividing the luminance image, such as gaussian filtering, weighted least squares filtering, guided filtering, bilateral filtering, and wavelet transformation. The luminance map of the image can be divided into high-frequency information lumadetail and low-frequency information lumabase by frequency division, wherein the high-frequency information lumadetail is obtained by subtracting or dividing the luminance luma of the image pixels from the filtered low-frequency information lumabase.
Step 2.2 mapping
And (3) mapping the image pixel points point by point according to the image related information obtained in the step (1). The corresponding output value Y can be calculated according to the mapping curve out . The specific content thereof can be represented by the following formula:
Y out ＝f(luma_base) (30)
wherein lumabase is the low-frequency brightness value of the image, f (luma_base) is the corresponding function of the brightness mapping curve, Y out And obtaining a mapping output value for lumabase according to the mapping curve.
MaxOut is the brightness range of the image after the image is reduced in bit width, Y out Temp is the mapped output value corresponding to the lumabase mapped to output bit width.
pix _ in is RAW image RAW data channel value, Y channel value of YUV image or each channel value of RGB image,for the brightness gain value, pix_out is the mapping value of the channel value corresponding to the reduced bit width, so that the color channel proportion of the original channel value after mapping is basically kept unchanged.
In the mapping process, it should be noted that the brightness gain corresponding to the current pixel point is obtained first in the mapping process, and the brightness value is obtained by combining the current pixel point and the peripheral pixel points, such as the brightness value obtaining method introduced in step 1. Therefore, the brightness value gain corresponding to the pixel and the brightness value gain corresponding to the peripheral pixels do not have larger difference, so that the image saturation can be ensured not to be changed greatly before and after mapping to a certain extent, and the problem of abnormal saturation caused by the high dynamic scene mapping process is avoided.
Step 2.3 high frequency back-addition
According to step 2.1 and step 2.2, the image is subjected to frequency division processing, then the low-frequency information and the high-frequency information of the image are acquired, and only the low-frequency information is subjected to mapping operation. Then, the high-frequency information is added back, and the high-frequency information is added back, so that the adding strategy is adjusted through a certain control parameter.
The high frequency back-adding coefficient r' may be a fixed constant or may be obtained by r (lumadetail) function according to the relationship between the high frequency data and the luminance value. The specific implementation of the r (lumadetail) function can perform a certain logic calculation for the difference value or the proportion value between the high-frequency data and the brightness value. For example, a certain threshold may be set by taking the absolute difference diff between lumadetail and luma as a function factor of the high-frequency back-adding coefficient, and different magnitudes of difference values correspond to different high-frequency back-adding coefficients, as shown in fig. 7. When the difference value is lower than diff_thr0, the high-frequency back-adding coefficient r ' is ratio_0, when the difference value is higher than diff_thr1, the high-frequency back-adding coefficient r ' is ratio_1, and when the difference value is lower than diff_thr0, the high-frequency back-adding coefficient r ' is interpolation between the ratio_0 and the ratio_1.
According to the calculated high-frequency back-adding coefficient, the final mapping output value calculation formula is as follows:
Pixel_out＝pix_out+lumadetail*r(lumadetail，luma) (33)
Wherein r (lumadetail) is a high-frequency back-adding coefficient, which can obtain a corresponding coefficient value through a back-adding coefficient lookup table, and then multiply the high-frequency information and then add back to pix_out to obtain a final mapping output value pixel_out.
The mapping output value pixel_out is the final optimization result of the automatic driving high dynamic scene image.
The embodiment of the invention provides an automatic driving high-dynamic scene image optimization system.
As shown in fig. 8, the autopilot high dynamic scene image optimization system provided in this embodiment may include:
the pixel brightness value acquisition module is used for acquiring the brightness value of each pixel in the automatic driving scene image;
the histogram data acquisition module is used for counting the brightness value of each pixel point in the image to obtain histogram data, and the histogram data comprises a brightness histogram and the number of pixels corresponding to the brightness interval;
the mapping curve acquisition module is used for judging the category of the automatic driving scene according to the histogram data and acquiring coordinates of the regulation points of the mapping curve to obtain a mapping curve;
and the mapping processing module is used for carrying out pixel-by-pixel mapping processing on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, so as to realize the dynamic range improvement on the automatic driving high dynamic scene image and finish the optimization on the automatic driving high dynamic scene image.
It should be noted that, the steps in the method provided by the present invention may be implemented by using corresponding modules in the system, and those skilled in the art may refer to a technical solution of the method to implement the composition of the system, that is, the embodiment in the method may be understood as a preferred embodiment for constructing the system, which is not described herein.
An embodiment of the present invention provides a computer terminal including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor, when executing the program, is operative to perform the method or operate the system of any of the foregoing embodiments of the present invention.
Optionally, a memory for storing a program; memory, which may include volatile memory (english) such as random-access memory (RAM), such as static random-access memory (SRAM), double data rate synchronous dynamic random-access memory (Double Data Rate Synchronous Dynamic Random Access Memory, DDR SDRAM), and the like; the memory may also include a non-volatile memory (English) such as a flash memory (English). The memory is used to store computer programs (e.g., application programs, functional modules, etc. that implement the methods described above), computer instructions, etc., which may be stored in one or more memories in a partitioned manner. And the above-described computer programs, computer instructions, data, etc. may be invoked by a processor.
The computer programs, computer instructions, etc. described above may be stored in one or more memories in partitions. And the above-described computer programs, computer instructions, data, etc. may be invoked by a processor.
And a processor for executing the computer program stored in the memory to implement the steps in the method or the modules of the system according to the above embodiments. Reference may be made in particular to the description of the previous method and system embodiments.
The processor and the memory may be separate structures or may be integrated structures that are integrated together. When the processor and the memory are separate structures, the memory and the processor may be connected by a bus coupling.
An embodiment of the present invention is also a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, is operative to perform the method of any of the above embodiments of the present invention, or to run the system of any of the above embodiments of the present invention.
According to the method, the system, the terminal and the medium for optimizing the automatic driving high dynamic range image, which are provided by the embodiment of the invention, the image mapping curve acquisition method adopts a multi-point control mapping curve and a self-adaptive dynamic range lifting mode, so that the regulation and control are flexible, the brightness of a low-brightness area of an image can be lifted, the high-brightness area is not influenced by other influences, the mapping method is flexible, and the self-adaptation is good; the high dynamic scene mapping method can effectively avoid the problem that the noise is excessively amplified in the mapping process of the image; the pixel brightness obtaining mode ensures that the brightness value of the current pixel keeps smaller difference with the gain of the peripheral pixels in the mapping process, the saturation is basically not affected, and the problem of abnormal saturation can be avoided in the image mapping process; considering the parameter values in multiple aspects, the method not only comprises the average brightness information value, the maximum brightness information value and the minimum brightness information value of the scene, but also comprises the discrete degree of the brightness histogram and the pixel number occupation ratio of each brightness interval, so that the classification and judgment of the image scene category can be more specific, different mapping curves are applicable to different scene categories, and the method is flexible and changeable.
Those skilled in the art will appreciate that the invention provides a system and its individual devices that can be implemented entirely by logic programming of method steps, in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc., in addition to the system and its individual devices being implemented in pure computer readable program code. Therefore, the system and various devices thereof provided by the present invention may be considered as a hardware component, and the devices included therein for implementing various functions may also be considered as structures within the hardware component; means for achieving the various functions may also be considered as being either a software module that implements the method or a structure within a hardware component.
The foregoing embodiments of the present invention are not all well known in the art.
The foregoing describes specific embodiments of the present invention. It is to be understood that the invention is not limited to the particular embodiments described above, and that various changes and modifications may be made by one skilled in the art within the scope of the claims without affecting the spirit of the invention.
Claims (9)
1. An automatic driving high dynamic scene image optimization method is characterized by comprising the following steps:
Acquiring a brightness value of each pixel point in an automatic driving scene image;
counting the brightness value of each pixel point in the image to obtain histogram data; the histogram data comprises a brightness histogram and the number of pixels corresponding to a brightness interval;
judging the category of the automatic driving scene according to the histogram data, and obtaining coordinates of the regulation points of the mapping curve to obtain the mapping curve;
according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, the image is mapped pixel by pixel, so that the dynamic range of the automatic driving high dynamic scene image is improved, and the optimization of the automatic driving high dynamic scene image is completed;
the mapping process of pixel-by-pixel is performed on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, and the mapping process comprises the following steps:
separating the high-frequency information in the brightness histogram to obtain low-frequency information in the brightness histogram;
based on the low-frequency information, mapping each pixel point of the image point by point according to a low brightness value, a high brightness value and a mapping curve of the whole image to obtain a mapping output value;
The separated high-frequency information is added back to the mapping output value, so that the dynamic range of the automatic driving high-dynamic scene image is improved;
the frequency division method for separating the high-frequency information in the brightness histogram comprises the following steps: gaussian filtering, weighted least squares filtering, guided filtering, bilateral filtering and wavelet transformation;
the mapping, based on the low-frequency information, each pixel point of the image according to the low brightness value, the high brightness value and the mapping curve of the whole image to obtain a mapping output value, including:
Y out ＝f(luma_base) (1)
wherein, luma_base is the image low-frequency brightness value, f (luma_base) is the function corresponding to the image low-frequency brightness mapping curve, Y out An output value obtained according to the mapping curve for the low-frequency information lumabase;
wherein MaxOut is the maximum value of brightness range after image bit width reduction, Y min Is of low brightness value, Y max Is a highlight value, Y out The temp is an output value corresponding to the mapping of the low-frequency information lumabase to the output bit width;
wherein pix _ in is the picture original data channel value,the pix_out is the output value after the bit width reduction corresponding to the channel value, namely the mapping output value;
said adding back the separated high frequency information to the mapped output value comprises:
Calculating a high-frequency back-adding coefficient;
and according to the high-frequency back-adding coefficient, carrying out back-adding on the separated high-frequency information to the mapping output value pix_out to obtain a final mapping output value pixel_out, wherein the final mapping output value pixel_out is as follows:
Pixel_out＝pix_out+lumadetail*r(lumadetail，luma) (4)
wherein r (luma) is a high-frequency back-adding coefficient, luma is high-frequency information, and luma is a pixel brightness value;
the high-frequency back-adding coefficient is obtained by adopting any one of the following modes:
setting a fixed constant;
and taking the absolute difference value of the high-frequency information and the brightness value of the pixel point as a high-frequency back-adding coefficient function factor, and obtaining a corresponding coefficient value through a back-adding coefficient lookup table according to a set difference value threshold.
2. The method for optimizing an autopilot high dynamic scene image of claim 1 wherein said obtaining a luminance value for each pixel in the autopilot scene image comprises:
according to the format type of the automatic driving scene image, a corresponding brightness value calculation method is adopted to obtain the brightness value of each pixel point in the automatic driving scene image; wherein:
the format type of the automatic driving scene image comprises: RAW, YUV, and RGB.
3. The method for optimizing an autopilot high dynamic scene image of claim 2 wherein,
For the automatic driving scene image format of the RAW, the brightness value calculating method includes any one of the following steps:
performing pixel-by-pixel original image statistics on an automatic driving scene image by adopting a block, wherein the maximum value, the average value or the average value of the maximum value and the minimum value of pixel values in the block with the current pixel point as a center and set pixel size is used as the brightness value of the current pixel point;
performing difference processing on a block with a set pixel size corresponding to each pixel point by using an interpolation method to obtain pseudo R ', G ' and B ' color values of the current pixel point; calculating the color values of the pseudo R ', G' and B ', and taking the maximum value, the average value of the maximum value and the minimum value or the weighted value of the color values of the pseudo R', G 'and B' as the brightness value of the current pixel point;
for YUV automatic driving scene image format, the brightness value calculation method comprises the following steps:
obtaining Y channel data corresponding to each pixel point, and calculating by adopting a block with a set pixel size taking the current pixel point as the center to obtain a corresponding brightness value;
for the RGB autopilot scene image format, the luminance value calculation method includes:
obtaining the brightness value of the current pixel point by adopting a weighting method;
And weighting the brightness value of the current pixel point and the brightness values of the peripheral pixel points by using a block with the set pixel size to obtain the final brightness value of the current pixel point.
4. The method for optimizing an image of an autopilot high dynamic scene according to claim 1, wherein said counting the luminance value of each pixel in said image to obtain histogram data comprises:
dividing the brightness range of the image into a plurality of brightness intervals;
carrying out brightness histogram statistics according to the brightness value of each pixel in the brightness interval;
and obtaining the number of pixels corresponding to each brightness interval according to the result of the brightness histogram statistics.
5. The method for optimizing an autopilot high dynamic scene image according to claim 1, wherein said determining an autopilot scene category from the histogram data and obtaining coordinates of a map curve control point, obtaining a map curve, comprises:
obtaining a low-brightness value, an average value and a high-brightness value of a current image according to the histogram data, wherein the low-brightness value represents a brightness value corresponding to a low-brightness pixel of the image, the high-brightness value represents a brightness value corresponding to a high-brightness pixel of the image, and the average value is the average value of brightness sums of all pixels in the image;
Analyzing a degree of dispersion of the luminance histogram, the degree of dispersion characterizing a dynamic range of the image;
counting the proportion of the number of pixels corresponding to the brightness interval to the total number of the image pixels;
judging whether the current image is a high dynamic scene image or not according to the ambient brightness value of the image, the difference value between the low brightness value and the high brightness value, the discrete degree of the brightness histogram and the proportion of the number of pixels corresponding to the brightness interval to the total number of pixels of the image; if yes, the control intensity of the mapping curve control points corresponding to the image is given, and a mapping curve is obtained; if not, dynamic range boosting is not required.
6. The automated driving high dynamic scenario image optimization method of claim 5, further comprising any one or more of:
the analyzing the discrete degree of the brightness histogram in the histogram data adopts an analysis of variance mode, and comprises the following steps:
wherein s is 2 Is variance, used to represent the degree of dispersion of luminance histogram, t is the number of luminance intervals of the input image, x [ i ]]For the luminance value corresponding to the ith luminance segment, hist [ i ]]The number of pixels corresponding to the ith brightness interval is n, and n is the total number of pixels of the image;
The method for judging whether the current image is a high dynamic scene image or not comprises the following steps:
and when the environment brightness value is more than or equal to a set environment brightness threshold value, the difference value is more than or equal to a set difference threshold value, the discrete degree is more than or equal to a set discrete threshold value and the proportion is more than or equal to a set proportion threshold value, judging that the image is a high dynamic scene image.
7. An autopilot high dynamic scene image optimization system for implementing the autopilot high dynamic scene image optimization method of claim 1, comprising:
the pixel brightness value acquisition module is used for acquiring the brightness value of each pixel in the automatic driving scene image;
the histogram data acquisition module is used for counting the brightness value of each pixel point in the image to obtain histogram data, and the histogram data comprises a brightness histogram and the number of pixels corresponding to a brightness interval;
the mapping curve acquisition module is used for judging the category of the automatic driving scene according to the histogram data and acquiring coordinates of the regulation points of the mapping curve to obtain the mapping curve;
and the mapping processing module is used for carrying out pixel-by-pixel mapping processing on the image according to the low brightness value, the high brightness value, the histogram data and the mapping curve of the whole image, so as to realize the dynamic range improvement of the automatic driving high dynamic scene image and finish the optimization of the automatic driving high dynamic scene image.
8. A computer terminal comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor is operable to perform the method of any of claims 1-6 when the program is executed.
9. A computer readable storage medium having stored thereon a computer program, which when executed by a processor is operable to perform the method of any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310024528.1A CN116152094B (en) | 2023-01-09 | 2023-01-09 | Automatic driving high dynamic scene image optimization method, system, terminal and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310024528.1A CN116152094B (en) | 2023-01-09 | 2023-01-09 | Automatic driving high dynamic scene image optimization method, system, terminal and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116152094A CN116152094A (en) | 2023-05-23 |
| CN116152094B true CN116152094B (en) | 2024-02-09 |
Family
ID=86359460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310024528.1A Active CN116152094B (en) | 2023-01-09 | 2023-01-09 | Automatic driving high dynamic scene image optimization method, system, terminal and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116152094B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103916669A (en) * | 2014-04-11 | 2014-07-09 | 浙江宇视科技有限公司 | High dynamic range image compression method and device |
| CN108109180A (en) * | 2017-12-12 | 2018-06-01 | 上海顺久电子科技有限公司 | The method and display device that a kind of high dynamic range images to input are handled |
| CN111683192A (en) * | 2020-06-11 | 2020-09-18 | 展讯通信（上海）有限公司 | Image processing method and related product |
| CN113096035A (en) * | 2021-03-31 | 2021-07-09 | 康佳集团股份有限公司 | High dynamic range image generation method and device, intelligent terminal and storage medium |
| CN113691739A (en) * | 2021-09-02 | 2021-11-23 | 锐芯微电子股份有限公司 | Image processing method and image processing device for high dynamic range image |
| CN115239578A (en) * | 2022-06-17 | 2022-10-25 | 展讯通信（上海）有限公司 | Image processing method and device, computer readable storage medium and terminal equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3295451B1 (en) * | 2015-05-12 | 2020-07-01 | Dolby Laboratories Licensing Corporation | Metadata filtering for display mapping for high dynamic range images |
-
2023
- 2023-01-09 CN CN202310024528.1A patent/CN116152094B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103916669A (en) * | 2014-04-11 | 2014-07-09 | 浙江宇视科技有限公司 | High dynamic range image compression method and device |
| CN108109180A (en) * | 2017-12-12 | 2018-06-01 | 上海顺久电子科技有限公司 | The method and display device that a kind of high dynamic range images to input are handled |
| CN111683192A (en) * | 2020-06-11 | 2020-09-18 | 展讯通信（上海）有限公司 | Image processing method and related product |
| CN113096035A (en) * | 2021-03-31 | 2021-07-09 | 康佳集团股份有限公司 | High dynamic range image generation method and device, intelligent terminal and storage medium |
| CN113691739A (en) * | 2021-09-02 | 2021-11-23 | 锐芯微电子股份有限公司 | Image processing method and image processing device for high dynamic range image |
| CN115239578A (en) * | 2022-06-17 | 2022-10-25 | 展讯通信（上海）有限公司 | Image processing method and device, computer readable storage medium and terminal equipment |
Non-Patent Citations (1)
| Title |
|---|
| 基于亮度自适应分段的高动态图像色调映射算法;刘颖;王倩;刘卫华;;电视技术(01);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116152094A (en) | 2023-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chang et al. | Automatic contrast-limited adaptive histogram equalization with dual gamma correction | |
| CA2563520C (en) | Methods and systems for converting images from low dynamic range to high dynamic range | |
| US8155465B2 (en) | Image processing method, medium, and system | |
| CN104200431A (en) | Processing method and processing device of image graying | |
| CN109584191B (en) | Self-adaptive image enhancement method and terminal based on histogram | |
| CN107820069B (en) | Video monitoring equipment ISP debugging method | |
| CN115115554B (en) | Image processing method and device based on enhanced image and computer equipment | |
| CN114331873B (en) | Non-uniform illumination color image correction method based on region division | |
| CN112116536A (en) | Low-illumination image enhancement method and system | |
| CN111127343B (en) | Histogram double-control infrared image contrast enhancement method | |
| CN112634384A (en) | Method and device for compressing high dynamic range image | |
| CN104639845A (en) | Generation method for high dynamic range image and device using method | |
| CN113449730A (en) | Image processing method, system, automatic walking device and readable storage medium | |
| Li et al. | Fast region-adaptive defogging and enhancement for outdoor images containing sky | |
| CN114173062B (en) | Image processing method, device, equipment and storage medium for image pickup equipment | |
| CN116152094B (en) | Automatic driving high dynamic scene image optimization method, system, terminal and medium | |
| CN114037641A (en) | Low-illumination image enhancement method, device, equipment and medium | |
| Albu et al. | One scan shadow compensation and visual enhancement of color images | |
| CN116645527A (en) | Image recognition method, system, electronic device and storage medium | |
| CN112381073A (en) | IQ (in-phase/quadrature) adjustment method and adjustment module based on AI (Artificial Intelligence) face detection | |
| CN113364994B (en) | Backlight compensation method and backlight compensation circuit | |
| CN113643651B (en) | Image enhancement method and device, computer equipment and storage medium | |
| CN115660997A (en) | Image data processing method and device and electronic equipment | |
| CN112348760A (en) | Image processing method and system | |
| CN115239578A (en) | Image processing method and device, computer readable storage medium and terminal equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |