WO2014114126A1 - 图像亮度调节方法和装置 - Google Patents
图像亮度调节方法和装置 Download PDFInfo
- Publication number
- WO2014114126A1 WO2014114126A1 PCT/CN2013/087193 CN2013087193W WO2014114126A1 WO 2014114126 A1 WO2014114126 A1 WO 2014114126A1 CN 2013087193 W CN2013087193 W CN 2013087193W WO 2014114126 A1 WO2014114126 A1 WO 2014114126A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- grayscale
- pixel
- gaussian
- processed
- 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.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
- G06T5/92—Dynamic range modification of images or parts thereof based on global image properties
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Definitions
- the invention relates to the field of digital and image processing, and particularly relates to a B-picture brightness two-section method and device. Background technique
- Digital Image Processing also known as computer image processing, converts image signals into digital signals and processes them using a computer.
- H image processing technology processes images to maximize the availability of useful information. Due to the influence of lighting, environment, and understanding, there will be national shadows and highlights in the circle image. How to perform the whole week on the shadow and highlights will directly affect the sensory feelings of human R during observation and H. Like quality.
- the automatic local adjustment of ffl like shadow and highlight is mainly based on the method of histogram, such as histogram equalization, that is, the non-fibre tt pull of the surrounding image, redistributing the image pixel value, so that a certain gray range
- histogram equalization that is, the non-fibre tt pull of the surrounding image
- redistributing the image pixel value so that a certain gray range
- the number of pixels in the interior is approximately the same, and the H-distribution of the given image changes the distribution of the histogram distribution.
- the implementation side provides an image brightness method and device.
- the technical solution is as follows:
- a method package for a picture brightness method Obtaining a to-be-processed image, taking a gray-scale image of each channel of the to-be-processed circle image to obtain a single-channel luminance pair, and a Gaussian filtering of the single-channel luminance image to obtain a Gaussian wave image r
- the stick shakes the gray-scale change rate of each pixel to process the B-image processing of the wide-distance to be processed, and obtains the processed a image;
- an image brightness two-segment device which is a device package «:
- a first acquiring module configured to acquire a to-be-processed circle image, and obtain a single-channel luminance image by using a gray level of each channel of the image to be processed;
- the second obtaining module obtains a Gaussian filtered image by performing a single channel luminance 3 ⁇ 4 image 3 ⁇ 4 row Gaussian filtering;
- the full-section module is used to dig the grayscale and pre-tilt side of the Gaussian filtered artifact, and adjust the grayscale of the Gaussian filtered image;
- a third obtaining module configured to adjust the gray-scale change rate of each pixel before and after the adjustment by using the adjusted Gaussian filtered image and the pre-adjusted Gaussian filtered H image;
- a fourth obtaining module configured to process the image of the to-be-processed circle aw ⁇ according to the gray-scale change rate of each pixel, and obtain the processed image by a;
- the tti module is used to output the processed w image ⁇
- the embodiment of the invention provides a method and a device for adjusting the brightness of the image, and obtains a single-channel brightness image according to the gray level of each channel of the surrounding image to be processed by acquiring the image to be processed;
- the Gaussian filtering image is acquired; according to the 3 ⁇ 4 Gaussian filtering circle, the grayscale ⁇ preset tt side, the gray level of the Gaussian filter group image of the section; 3 ⁇ 4 section!
- Gaussian fiber S image and pre-section The Gaussian filter enclosing image obtains the gray-scale change rate of each pixel before and after the adjustment; the image is processed according to the gray-scale change rate of each pixel, and the processed image is obtained; The processed image is used to process the gray scale of each channel of the image.
- the technical solution provided by the present invention fully utilizes the entire color channel table 3 ⁇ 4 capability, and ensures the circle by filtering the single pass luminance image by 3 ⁇ 4 Gaussian filtering.
- the high-filtering artifacts are adjusted according to the preset two, so that the shadow and highlight E-domains in the Gaussian filtered image tend to be neutral, which is more suitable for human observation.
- FIG. 1 is a flow chart of an image brightness adjustment method provided by the implementation of the present invention.
- FIG. 2 is a flow chart S of an image brightness adjustment method provided in an embodiment of the present invention ;
- H 3 is a block diagram of the H-image brightness adjusting device provided in the implementation side of the present invention
- the circle 4 is a block diagram of the adjusting module 303 in the embodiment of the present invention
- ⁇ 5 is the fourth acquisition module 305 in the implementation of the present invention.
- FIG. 1 is a flowchart of a method for adjusting the brightness of an image according to the implementation of the present invention.
- the execution body of the implementation is considered to have a device for processing a circle image, such as a personal computer, a camera device or a service, etc., see H i, Methods include:
- the image to be processed is a digital a image, that is, a two-dimensional array represents a curved image.
- the basic element of the digital H image is a pixel (or pixel, Pixel), which is obtained by sensitizing the continuous space in the digitization of the analog image.
- Each pixel has an integer row (high) and an integer column (width) position 3 ⁇ 4, and each pixel has an integer gray value or an integer color value.
- the gradation is a quantization of the brightness change of the image, and the depth of the brightness is not expressed by I, and the gradation is usually 256 gradations, that is, the gradation of the gradation is 0 ⁇ 255, and the change of 0-255 represents The brightness is from deep to light, and the color of the corresponding image is from ⁇ to white.
- the image W grayscale range is also expressed as 0 1, 0 for ⁇ leg color, 1 for white, 04 for the number of children indicating dark to light gray.
- the single-channel brightness circle image is an H image composed of JS large gray scale obtained by ft the gray scale of each channel of each image point in the image to be processed.
- the image has three channels as a single channel luminance H image pair (R ( Red, Red), G (Green, Green XB (Blue, It))
- Each channel of the three channels) image has I: ....i gray scale, each gray scale corresponds to each channel, then for each pixel point, three pixel gray for the pixel point
- the degree of the fiber line tt is obtained, and the maximum value among the three gray levels is obtained, and the maximum value is taken as the gray level of the pixel point W to obtain a single channel brightness circle image.
- the gray scale of each pixel is obtained from the gray scale of R, G, and B of each pixel of the image to be processed, and then each will be The maximum gradation of the pixel is used as the gradation of each pixel of the single-pass ⁇ luminance circle image, thereby acquiring a single-channel luminance image.
- the image can be converted into an RGB image, and then the single channel image can be obtained. ill can directly use the image of the channel used to represent the brightness in other color mode images as a single image. 3 ⁇ 4 way brightness image.
- the image of the image is RGB E, and the image is the edge of the fi 1
- the Gaussian filter is used to filter the single channel luminance H image.
- the Gaussian filter radius is 5 20 pixels.
- the ffl image to be processed is convoluted with the filter function of the Gaussian filter to obtain a Gaussian filtered image.
- the Gaussian filter image size must be kept at a 3 ⁇ 4 size to the H image to be processed.
- a single channel luminance image is input into a Gaussian filter, and a Gaussian filter is applied to filter the Gaussian filtered image to obtain a Gaussian filtered image.
- ⁇ is the Gaussian distribution parameter, which determines the smoothness ⁇ of the ffl image after Gaussian filtering is larger, the smoothness of the image is higher, ⁇ The smaller the image, the lower the smoothness of the image.
- 3 ⁇ 4A For 2D image processing, you can use two-dimensional Gaussian filter metrics to smooth the surrounding image 3 ⁇ 4 lines, or you can use a two-dimensional Gaussian filter function to smooth the circle image. In the process, the 2D Gaussian filter is often used to achieve smooth filtering of the image.
- ' represents the two-dimensional Gaussian filter "position information
- j is the Gaussian distribution parameter
- the Gaussian filter is a class-based weighting function based on the shape of the Gaussian function.
- the slit smoothing filter is Gaussian filtering, and the Gaussian filter is used to perform Gaussian filtering on the single-channel luminance image.
- the size of the root image and the degree of filtering are required to adjust the number of filter capsules in the filtrate * and the respective parameters of the number of 3 ⁇ 4, and the implementation of the present invention does not specifically determine this.
- the Gaussian filter ffl of the predicate is like a male.
- the number of pixels corresponding to each gray level can be obtained, and the gray level of the preset proportional pair is used as the first gray level, which is smaller than the first
- the area composed of the pixels of the gray scale is used as the garden area, and the pixel points larger than the first gray level are composed of the Weng area as the highlight area, and after obtaining the Gaussian filtered image, the area is called the shadow area.
- the gradation of the K-field and the highlight region ⁇ pixel points, and the adjusted gradation is obtained.
- the gray level change rate of each pixel point can be obtained by comparing the gray level of each of the adjusted pixel points of the adjusted Gaussian filter H image and the Gaussian filter image before the predicate.
- the grayscale change rate of each pixel is used for the adjustment of the brightness of the image to be processed.
- the gray level of each pixel of the Gaussian filter image after the node is compared with the gray level of the Gaussian filter image of the pair before the predicate, and the implementation of the present invention is required.
- the two of the two are the same point of the image.
- the ratio of the gray level of each pixel can be obtained by the ratio of the gray level of each pixel of the magic.
- p is the index of the image
- the value of p is 3 ⁇ 4, MxN]
- M is the number of image lines
- N is the number of surrounding images
- M is a positive integer
- £ 2 ⁇ is Gaussian after the predicate Filter the bit of the image Set the gray level of the pixel point of the coordinate p
- I ⁇ is the gray level of the pixel point whose bit coordinate of the Gaussian filtered image is p.
- the pixel belongs to the R shadow region, and the larger the value of the change rate is, the larger the luminance i of the pixel is.
- the gray-scale change rate is equal to 1, the pixel It is neither a T-shaded area nor a high-light E-domain, and the pixel is immersed in what is changed.
- the gray-scale change rate is less than 1, the pixel belongs to the highlight area, and the value of the change rate is smaller, the pixel point The greater the brightness adjustment ratio.
- the gray value of the WK shadow region in the image to be processed can be adjusted. High (ie, brighten the shadow area), and lower the gray value of the highlight area (ie, darken the highlight area), and use the image as the processed group image
- a single-channel brightness surrounding image is obtained according to the gray scale of each channel of the captured image; the Gaussian filtering of the single-channel brightness circle image is performed to obtain the high-body filtering circle.
- the root-shake pre-sequence is full of Gaussian filter-sensitive images, which makes the shadow and highlight areas of the Gaussian filtered image tend to be neutral, which is more suitable for human obscurity observation.
- 2 is a specific flow circle of an H-image brightness adjustment method according to an embodiment of the present invention. In this embodiment, only the image to be processed is an RGB H image as an example. The following is mainly combined with the specific implementation of step 103 and step 105 in FIG.
- MBh determines the number of W-pixel points for each gray scale in descending order of gray scale according to the gray scale of each pixel of the Gaussian filtered image.
- the number of pixels of the Gaussian filter image and the number of pixels of the image to be processed need to be consistent, and the gray rate of change obtained by the high-interpolation image in the subsequent step is used to process the image to be processed. It is necessary to make the image of the Gaussian annihilation image the same as the total number of pixels of the group to be processed. If the size of the image to be processed is a row, the number of pixels of the image to be processed is ⁇ , and the number of pixels of the Gaussian filter image should be MxN, and the number of pixels of the pair of H images is "processing". The filtering process of step 202 is implemented.
- the first number is obtained by multiplying the number of pixels of the Gaussian filter image by the multi-input ratio, and the first value can be used to obtain the first-gradation of the T-description, and the brilliant can distinguish the WW shadow region and the circle in the circle image.
- the light area ⁇ can be compared with the actual image of the circle image. If the area of the light shadow in the S image is large, the highlight area is small, if the ratio of the town preset is 3 ⁇ 4 « People, such as 80%, if the shadow area in the image is small and the highlight area is large, you can set the pre-renderer to a smaller song, such as 203 ⁇ 4.
- the first gray level is obtained according to the first value and the determined number of pixels of the determined gray scale.
- the first gray scale is used to distinguish the shadow area and the highlight area of the H image.
- the pixel is smaller than the first gray pixel, and the pixel point larger than the first gray is used as the highlight region, so as to adjust the brightness of the two E-domain lines. .
- the number of pixels corresponding to each gray if of the Gaussian filtered image is accumulated according to the order of gray scale from small to large, until the previous accumulated child is smaller than the first value, and the accumulated value of the second time is equal to the first When the value is counted, stop the It accumulate, and the corresponding gray level on the right circle is used as the first gray level.
- the step 1031 1033 is a process of acquiring the number of pixels of the Gaussian chopping image, the gray level of each pixel, and the preset ratio, and acquiring the first gray scale for distinguishing the W shadow region and the highlight region of the H image.
- the area composed of the first pixel of the wide path is a light shadow area.
- Pixel brain grayscale where p is the position of the first pixel of the Gaussian subtraction image ⁇ , 2 ( ) is the gradation of the first pixel after the festival, and the gradation of the first pixel before the adjustment is pre- Set parameters,
- HI is the first gray scale
- the first pixel point refers to any pixel point in the high-definition wave image
- used by the above formula ⁇ 3 ⁇ 4 represents the bit of the first-pixel point.
- W coordinate if the size of the Gaussian filtered image is M ⁇ N column , the value of p is 3 ⁇ 4B is "1, MxN] 0.
- the position coordinates of the two-dimensional image it can be represented by two component data, or can be expressed by -, for example, for an image of size M ⁇ N column
- the position of the pixel in the row/column of the row can be expressed as /], or can be expressed as Mxi'+j'.
- the two representations can be converted to each other. In the implementation of the present invention, only one number is used.
- the position coordinates of the pixel are as shown in the example a. In the actual application, one data or two component data may be used, and the implementation side of the hairpin does not specifically limit this.
- the above formula is used to traverse the high-guess filtering H image in which the gray level is smaller than the first gray level pixel point, and the T-to-picture area adjustment is realized.
- the non-money tt transformation formula on the pixel in the image ttPJ shadow region After the nonlinear adjustment of the gray level of the point, the brightness of the shadow area is increased, the local contrast is enhanced, and the detail information is prominent, which is convenient for human eyes to observe.
- the gradation of the first pixel of the Gaussian filtered image is equal to the first gradation, the gradation of the first pixel is not adjusted.
- the gray level of the first pixel point is suitable for the observation of the human eye, and the brightness is moderate, so the gray levels of the pixel points are not adjusted.
- ⁇ (p) is the gradation of the adjusted first pixel, and is the gradation of the first pixel before adjustment.
- the first pixel of the Gaussian filtered image When the gradation of the first pixel of the Gaussian filtered image is greater than the first gradation, then the first pixel
- the above formula is used to traverse the pixels of the Gaussian filtered image whose gray level is larger than the first gray level, and the adjustment of the highlight area is realized.
- Steps 1031-1034 are an example of a specific process of adjusting the gradation of the Gaussian filtered image according to the gradation and the preset ratio of the Gaussian filtered image in the above step 103.
- step 105 The specific process of step 105 is described below.
- 1051 After obtaining the gray level change rate of each pixel before and after the adjustment, 1051: multiplying the gray level change rate of each pixel point by the gray level of the corresponding pixel point in the image to be processed to obtain a second Grayscale.
- the image to be processed of tlj is an RGB network image. Therefore, the gray levels of the three channels of R, G, and B of the image to be processed are separately processed, and the gray rate of change of each pixel and the pixel corresponding to each channel W are Multiply the grayscale 3 ⁇ 4 lines to get 3 ⁇ 4 of the second gray level per pixel per pass'
- the setting of the second preset value needs to be based on the a-like value of the line.
- unsigned 8-bit integer uint8
- the second pre-value can be set to 255, and when the second gradation of one pixel is greater than 255, 255 is used as the pixel.
- Grayscale Use the double precision (that is, when the W image is grayscale, the second preset value 3 ⁇ 4! is set to 1, 3 ⁇ 4, a pixel is clicked, and the second grayscale is greater than 1, 1 is taken as the ft-pixel.
- the gray scale has a large range of gray scale change rate.
- the second gray scale may be exceeded beyond the gray scale of the image display. In practical applications, it is required to be based on the specific numerical class. Second, whoever is worth the value, therefore, the implementation of the present invention does not limit the size of the second preset value.
- the circle image gray scale is rounded, so that the gray scale bacteria surrounding the image is kept within 3 ⁇ 4 of the circle image display.
- Steps 1051 to 1053 are an example of the specific process of processing the processed image a-line processing according to the gray-scale change rate of each pixel according to the above-mentioned step 105, and obtaining the processed image.
- the 3 ⁇ 4 circle is disposed only by multiplying the gray scale change rate of each pixel point and the gray level of the corresponding pixel point in the image to be processed. Like a line of processing. After the image to be processed is processed through a series of processes, the PJ shadow area and the highlight area are adjusted, and the adjusted I»H image s is combined with the observation of the human eye and the acquisition of detailed information.
- the implementation side of the present invention obtains a single-channel luminance image according to each of the overnight grayscales of the to-be-processed circle image by acquiring a to-be-processed image; obtaining a Gaussian filtered image by performing Gaussian filtering on the single-channel luminance image; The gray scale of the Gaussian filtered image and the pre-t-down, the gray level of the Gaussian filtered image built by the i-section; the adjusted Gaussian-off-wave image and the pre-joint Gaussian filtered image, The gray-scale change rate of the pixel; the image to be processed is multiplied by multiplying the gray-scale change rate of each of the pixel points with the gray-magic of the image of the monster in the captured B image Line processing; output the image after the capsule, to the gray level processing of each channel of the image, the technical solution provided by the BJ implementation of the present invention fully utilizes the performance of the entire color channel, through the image of the dry channel brightness Gaussian filtering is performed to
- FIG. 3 is a connection frame of a tffl image brightness adjusting device provided in the implementation of the present invention, and the device includes:
- the first obtaining module 301 is configured to acquire a to-be-processed image, and the root-drilling image is captured by the ffl image of each channel to obtain a single-channel brightness image.
- the first acquiring module 301 is configured to obtain an image to be processed, compare the gradations of each bright channel of each pixel, and obtain a maximum value of each pixel gradation, and use the maximum value as The gray level of each pixel is captured to obtain a single channel luminance image.
- the second obtaining module 302 is configured to perform Gaussian filtering on the single-channel luminance image to obtain a muse filter image.
- the predicate module 303 is configured to adjust the Gaussian filtering and the gray level of the image according to the gray scale and the preset ratio of the Gaussian filtered image.
- the adjustment module 303 includes:
- the first gradation acquires the dry element 3031, and obtains the number of the pixel points of the Gaussian filtered ffl image, the gradation of each pixel point, and the pre-ratio, to obtain the difference between the bright shadow area and the highlight area for distinguishing the image.
- a gray scale A gray scale. .
- the first grayscale acquisition Pane element is used to determine the number of W pixel points corresponding to each gray level according to the gray level of each pixel point of the Gaussian filter image, according to the gray level from small to large.
- the number of pixels of the Gaussian filter image and the preset ratio determine a first value; and the first value is obtained by summing the first value and the determined number of the wide gray pixel points of the respective gray levels to obtain the first gray level.
- the first side unit 3032 is used to calculate the gray level of each pixel according to the Gaussian filtered image.
- the first gray scale is described, and the Gaussian created by the shoulder section filters out the gray scale of the B image.
- the gray level of the first pixel is smaller than the first gray level, and the area composed of the first pixel is the ⁇ shadow area, and the application is £ 2. (,) () (i
- the leg S adjusts the gradation of the first pixel; when the gradation of the first pixel of the Gaussian filtered H image is equal to the first gradation, the gray of the first pixel is not a row; 3 ⁇ 4 Gaussian filtering The gray level of the nil image is greater than the first pixel, and the E field of the first pixel is captured as a highlight region, and the application is a high-energy filter circle image.
- the first pixel point W position coordinate, £ 2 ( ) is the gray level of the first pixel after adjustment, £ (
- the third obtaining module 304 is configured to compare the Gaussian filtering ffl image after the 3 ⁇ 4 section and the Gaussian filtering circle image before the singular section, and obtain the grayscale change rate of each pixel before and after the adjustment.
- the fourth obtaining module 305 is configured to process the to-be-processed image according to a gray-scale change rate of each pixel created, and obtain a ⁇ : the processed circle image.
- the fourth obtaining module 305 includes a second grayscale acquiring unit 3051, configured to compare the grayscale change rate of each pixel point with the image to be processed Multiplying the pixel points by the gray level to obtain a second gray level;
- a second adjusting unit 3052 wherein the second gray level of any pixel point is greater than the second preset child time, and the gray side section of any pixel point is set to a second preset value;
- the S image acquisition unit 3053 uses ⁇ to acquire the adjusted image as the processed artifact.
- the output module 306 is configured to output the processed circle.
- the first acquisition module 301 of the embodiment of the present invention acquires a to-be-processed image, and acquires a single-channel 3 ⁇ 4 degree image according to the gray scale of each of the captured circle images W; the second acquisition module 302 pairs the single Channel luminance image 3 ⁇ 4 row Gaussian filtering, obtaining Gaussian filtering surrounding image; adjusting module 303 according to the grayscale ⁇ preset ratio of the Gaussian filtering ffl image, "High-filtering image W grayscale; third acquisition module 304 comparison After the capsule section height « «-wave pattern and i-chamber section Gaussian filter ffl image, the gray-scale change rate of each pixel before and after adjustment is obtained; the fourth acquisition module 305 is based on the gray-scale change rate of each pixel point.
- the captured circle is processed to obtain the processed image; the output module 306 outputs the processed image by a to process the grayscale of each image, and the method provided by the embodiment of the present invention is fully Utilizing the entire color channel expression ability, the Gaussian filtering of the single-channel luminance image is used to ensure the continuity of the image.
- the a-stick is compared with the side-by-week Gaussian filter circle image to make the shadow in the Gaussian filtered image.
- the highlights and highlights tend to be neutral, making them more suitable for human eye » observation.
- the above-mentioned implementation side provides the image brightness adjustment device to process the ⁇ shadow and the high area of the circle image, and only the division line of each functional module of h is mentioned.
- ⁇ is to be based on ffi
- the function assignment is completed by different w function modules, and the internal internals are divided into different power cage modules, which are stored in the computing device, and are at least one processor in the computing device, ie, t3 ⁇ 4 h Describe all or part of the function.
- the implementation of the image brightness adjustment device s and the image brightness adjustment method are the same as the embodiment, and the law implementation process is shown in the method embodiment, and is not praised here.
- the storage medium mentioned above may be a read only memory, a magnetic disk or an optical disk or the like.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
提供了一种图像亮度调节方法和装置。该方法包括:获取待处理图像,根据所述待处理图像的各个通道的灰度获取单通道亮度图像;对所述单通道亮度图像进行高斯滤波,获取高斯滤波图像;调节所述高斯滤波图像的灰度;比较调节后的高斯滤波图像和调节前的高斯滤波图像,获取调节前后每个像素点的灰度变化率;对所述待处理图像进行处理获取经过处理后的图像;输出经过处理后的图像。通过对图像各个通道的灰度的处理,充分利用了整个颜色通道的表达能力,通过对单通道亮度图像进行高斯滤波,保证了图像整体的连续性,通过根据预设比例调节高斯滤波图像,使高斯滤波图像中阴影和高光区域趋于中性,更适合于人眼的观察。
Description
说 明 书 圈像亮度调节方法和装置 技术领域
本发明涉及数字,像处理领域, 特 涉及一种 B像亮度俩节方法和装 。 背景技术
数字图像处理(Digital Image Processing, DIP)又称为计算机 H像处理, 它 是 将图像信号转換成数字信号并利用计算机对其进行处理 过程。 通过数字
H像处理技术对图像迸行处理, 最大限度地获取有用信息。 而由于光照、 环境、 懂备等因素的影响, 圈像中会存在國影和高光部分, 如何对明影和高光 部分迸行 i周整, 会直接影响人 R在观察时的感官感受以及 H像 质量。
对 ffl像闘影和高光的自动局部调整主要采用基予直方图的方法, 如直方图 均衡化, 即对圍像纖行非纖 tt拉傅, 重新分配團像像素值, 使一定灰度范围内 的像素数量大致相同, 把给定圏像的直方 H分布改变虐" 勻"分布 W直方图分 布。
a过直方图的方法调整图像的亮度信息, 由于直方图均衡化对处理 图像 数据不加¾择, 因此会增加背景嗓声的对比度并且 »低有用信号的对比度, 面 i处理后 I»图像容易出现局部不连续的情况, 进而会导致丢失图像屮的细节信 息。 发明内容
为了解决圈像亮度谓节问题, 本发嘆实施侧提供了一种 a像亮度 w节方法 和装置。 所述技术方案如下:
一方面, 提供了一种 a像亮度谓节方法 所述方法包
获取待处理图像, 根攝所述待处理圈像的各个通道的灰度获取单通道亮度 对所 单通道亮度图像 ¾行高斯滤波, 获取高斯德波图像 r
根据所述高 if滤波图像翁灰度和预殺 例, 调节所述高斯滤被 a像的灰度; 比较调节后的高斯滤波图像和调节前的高斯滤滅 H像, 获取调节前后每个 像素点的灰度变化率;
棍摇所述每个像素点的灰度变化率对廣途待处理 B像纖行处理, 获取经过 处理后的 a像;
输 Λ经过处理后翁图條„
另一方面, 提供了一种图像亮度俩节装置, 所¾装置包 «:
第一获取模块, 用于获取待处理圈像, 根揚所述待处理图像的各个通道的 灰度获取单通道亮度图像;
第二获取模块, 予对所建单通道亮度 ¾像¾行高斯滤波, 获取高斯滤波 图像;
满节模块, 用于根掘 述高斯滤被圏像的灰度和预 ¾tfc側, 调节所 ¾高斯 滤滅图像的灰度;
第三获取模块, 用于 ft较调节后的高斯滤波图像和调节前》高斯滤波 H像, 获取调节前后每个像素点的灰度变化率;
第四获取模块, 用于根据 逑每个像素点的灰度变化率对 ¾待处理圈像 aw亍处理, 获取经 a处理 w图像;
输 tti模块, 用于输出经过处理后 w图像 β
本发明实.施例提供的技术方案带来》有益效果是
本发明实施例提供了一种 ffl像亮度调节方法和装置, 通过获取待处理图像, 极据所建待处理圍像曲各个通道的灰度获取单通道亮度圍像; 对所述单通道亮 度圈像进行高斯滤波, 获取高斯滤波圍像; 根据所 ¾高斯滤波圈像翁灰度 Λ预 设 tt侧, 节所途高斯滤波團像的灰度; 较 ¾节!!的高斯纖波 S像和谓节前
曲高斯滤波圍像, 获取调节前后每个像素点翁灰度变化率; 植据 逾每个像素 点的灰度变化率对所述待处理图像进行处理, 获取经过处理后的圍像; 输出经 过处理后的圍像來对图像各个通道的灰度的处理, 本发明实施倒提供的技术方 案充分地¾用了整个顔色通道 表¾能力, 通过对单通通亮度圍像 ¾ 高斯滤 波, 保证了圈像整体的连续性, 邇过植据预设 俩调节高 滤波圏像, 使得高 斯滤波图像中阴影和高光 E域趋于中性, 以更适合于人 观察。 附图说明
为了夏清楚地 ¾明本发噴实施倒中的技术方案, 下面将对实施例描 ¾中 需要使用的耐图作简单地介绍, 显而易见地, 下面摇建中的附團仅 是本发 的一些实施俩, 对于本领域普通技术人员来讲, 在不付出创造性養动的前提下, ¾ 以根据这些 »图获得其他的 S。
图 1是本发明实施倒中提供的一种图像亮度调节方法流程團;
图 2是本发明实旌例中提供的 』种图像亮度调节方法流程 S;
H 3是本发嚷实施侧中提供翁一种 H像亮度调节装置的结 框图; 圈 4是本发 实施例中的调节模块 303的蛣构框图;
圏 5是本发明实施俩中的第四获取模块 305 W结枸框 Β 具体实施方式
为使本发嚷的 的、 技术方案和优点更加清楚, 下面将结合附图对本发明 实施方式作«—歩地详細描逑 Q
图 1 是本发 实施倒 提供 一种图像亮度调节方法流程圈, 该实施倒的 执行主体 以为具有圈像处理能力曲设备, 如个人†算机、 摄像设备或服务》 等, 参见 H i, 该方法包括:
1015 获取待处理图像, 揋据所逑待处理 H像的各个通道的灰度获取单通道 亮度■像
在一个示侧中, 待处理圍像为数字 a像, 即由二维数组形 表示曲图像。 本領域技术人员可以获知, 数字 H像的基本 素是像素(或像元, Pixel), 像索 是在模拟图像数字化时对连续空间纖行离敏化得到的。 每个像素具有整数行 (高) 和整数列 (宽) 位置 ¾标, 同时每个像素都具有整数灰度值或整数顏色 值。
其中, 灰度是对图像亮度变化的一个量化, 用 I表不亮度的深浅, 通常灰 度量化为 256个灰度级, 即灰度的舊围为 0·255, 该 0-255的变化表示了亮度的 从深到浅, 对应圏像中 W顏色为从黒到白。 图像 W灰度范围还 表示为 0 1, 0表示 ·腿色, 1表示白色, 04之闻 数僮表示从深到浅的灰色。
其中 , 单通道亮度圈像是 通过 ft较待处理图像中每个像 点的各个通道 的灰度获取的 JS大灰度组成的 H像。
对亍只有一个邇道的圈像如灰度圈像, 将该图像作为单邇道亮度 H像 对 具有三个通道 ( R ( Red , 红色)、 G (Green, 绿色 X B ( Blue, It色) 三个通道) 的1!像》 图像 每个像素点均 I有:....i个灰度, 每个灰度对应 个通 道, 则针对每个像素点, 对该像素点 三个通道 灰度纖行 tt较, 获取三个灰 度中的最大值, 将该最大值作为懷像素点 W灰度, 得到单通道亮度圈像。
也就是说, 对于具有三个通道的特处理■像, 较该待处理圏像的每个像 素点的 R、 G、 B的灰度 获取每个像 点的最大的灰度, 之后将每个像素点的 最大的灰度作为单通逍亮度圈像的每个像素点的灰度, 从而获取单通道亮度图 像。
而对于其他顔色模式的图像, 可 先将圏像转化为 RGB圍像, 再获取其单 通道 ^度 像, ill可 直接将其他顔色模式图像中的用于表示亮度的通道所对 的图像作为单 ¾道亮度圍像。 在本发明实施例中, 仅¾待处现图偉为 RGB E 像为侧进 fi1说明
102: 对所述单通道亮度■像进行高斯滤滅, 获取高斯滤波 11像》
在一个示例中, 在通过上逮歩骤 101 获取单通道亮度图像之后, 为了保持
圈像的连续性, 需要对 B像进行滤波, 由于高斯滤波器具有较好的滤波效果, 所以, 优迪地, 在本发明实施俩 Ψ采用高斯滤波器对单通道亮度 H像鍾行滤減。 同时, 根据圈像的大小, 逢择適合的尺寸, 优逢地, 高斯滤波半径为 5 20个像 素。
通过将待处理圏像输入 ft斯滤波器, 使得待处理 ffl像与高斯滤波器的滤波 函数纖行卷积 算, 获取高斯滤波图像。 需要说嚷 是, 高斯滤波圏像 大小 须和待处理 H像大小保持一 ¾。 通过对单遣道图像逮行高斯滤波, 获取平滑翁 高斯浦被图像, 避兔了 H像灰度翁不连续。
在该歩骤 102 中, 将单通道亮度圏像输入高斯滤波《, 应用高斯滤波》» 滤波函数对其进行高斯滤波, 获取高斯滤波图像。
优進地, 该歩骤 102可应用下式 ( 1 ) 中厨示的一维零均值离散高斯滤被器 函数进行《:断滤波: g(x) = i^ ( 1 )
其中, 是 ( 斯滤波器 位置坐标, ( 是高斯滤波器在 处的数僮, σ是 高斯分布参数,决定了高斯滤波后 ffl像的平滑度 σ越大,圏像 平滑度就越高, σ越小, 图像的平滑度就越低。
¾A» 对于二维圏像处理来 ¾, 可 使用两次一维高斯滤波涵数对圍像 ¾行平滑滤波, 也可 使用一次二维高斯滤波函数对圈像迸行平滑滤波 在实 ϋ使駕过程中, 常采用二维高斯滤¾¾数》 ― - -次性实现对图像的平滑滤波。
优逾地, 该歩骤 102可应用下 (2)中所示的二维零均僮离散高斯滤波器 函数迸行高斯滤波: g[i, j] = e 2σ ( 2)
其中, '、 代表二维高斯滤波器《位置信息, 是二维高斯滤波器在位 置 值, j是高斯分布参数, 决定 T二维高斯滤波后圈像的平滑度, σ越
大, 图像 »平滑度就越高, σ趋小, 图像的平滑度就越低。
在本发明实施侧 ÷ , 高斯滤波》是 - 类根据高斯函数的形状来逾择权值 W 缝性平滑滤敏器, 采用高斯滤波器对所述单通道亮度图像进行高斯滤波, 可以
¾到平滑图像、 «Ρ余噪声 |%效果, 避兔了圈億 W不连续性
需要说明》是, 在实际 用过程中, 需要根攝 Η像的大小和滤波程度调节 滤液 *中的滤波囊数以及 ¾数的各个参数, 本发明实施倒对此不作具体 R定。
103: 根据所述高斯滤波圈像的灰度和预设比例, 谓节所逑高斯滤波 ffl像 雄。
具体而言, 通过对上述歩骤 102获取的高斯滤波圍像翁灰度 行统计, 可 以得到各个灰度对应 像素点 个数, 将预设比例对 的灰度作为第一灰度, 小于第一灰度的像素点組成的区域作为園 域, 而将大于第一灰度的像素点 組成翁区域作为高光区域 在获取高斯滤浓圏像购國影 £域 高 A区域后, 分 别谓节園影 K域和高光区域 Φ像素点的灰度, 获取调节后 灰度。
104: 比较谓节后的高斯滤被圏像和调节前的高斯減被图像, 获取调节前后 每个像素点的灰度变化率。 ·
具体而言, 每个像素点的灰度变化率可通过将调节后的高斯滤波 H像和谓 节前的高斯滤波围像的每个对 曲像素点的灰度做比値遞算来得到, 每个像素 点的灰度变化率用于待处理围像 W亮度的调节。
在一个示例中, 将谓节后的高斯滤波围像的各个像素点的灰度和谓节前的 对 的高斯滤波图像的 个條素点翁灰度做比值运算, 需要 的是, 本发明 实施俩中的对彦擔 是位置相同的像尜点, 通过对魔的每个像素点的灰度的比 值, 可 获取每个像素点的灰度变化率 该灰度变化率†算公式如下
其中, p是图像的倥置 标, p的取值 ¾围为 , MxN], M表示图像 行 数, N表示圍像翁列数, M、 为正整数, £2ω为谓节后的高斯滤波圏像的位
置坐标为 p的像素点对産的灰度, I ^ 为高斯滤波图像的位 坐标为 p的像素 点对应 灰度。
当灰度变化率大亍 1 时, 该像素点属于 R影区域, 变化率的值越大, 该像 素点的亮度 i周节比例越大; 当灰度变化率等亍 1 时, 该像素点既不属 T阴影区 域也不属于高光 E域, 该像素点浸有做 何》变化; 当灰度变化率小于 1 时, 该像素点属予高光区域, 变化率的值越小, 该像素点的亮度调节比例越大。
105: 根据 途每个像素点的灰度变化率对厨 ¾待处賴圈像 a行处理, 获取 经过处理 的8像《
具体丽言, 通过将每个像素点的灰度变化率和待处理 H像中对魔 W每个像 素点 灰度通行乘积 ¾算, 可以将待处理圉像中 WK影区域的灰度值调高 (即, 将明影区域调亮), 而将高光区域的灰度值调低 (即, 将高光区域懂暗), 将该 图像作为经过处理后的團像
106: 输出经过处理后的图像
将经过处理后的圏像输出, 与原图像相比, 该 H像爝加了圏像的对比度, 具有较多的細节信息, 适合于入眼的观察 a
本发明实施例通过获取待处理團像, 裉据所逮待处理圍像的各个通道翁灰 度获取单通道亮度圍像; 对所建单通道亮度圈像进行高斯滤波, 获取高身 Ϊ滤波 圈像; 根据所述高斯滤波 ffl像的灰度和预设比例, 调节廣述高衛滤波顯像》灰 度; tt较囊节后的高斯滤 ¾图像靡调节前的高斯滤波图像, 获取调节前后每个 像素点的灰度变化率; 极据所述每个像素点的灰度变化率对所述待处理图像 a 行处理, 获取经 ^处理后的围像; 输出经过处理后 WH像来对图像各个通道 灰度的处理, 本发 实施例提供的技木方案充分地利用了整个颜 fea道的表达 能力, 通过对单通道亮度图像 «行高斯滤波, 保证了 a像整体的连续性, 通过 根摇预缓 例满节高斯滤敏圍像, 使高斯滤波图像中明影和高光区域趋予中性, 更适合于人隱翁观察。
图 2是本发明实施例提供的一种 H像亮度调节方法的具体流程圈, 在本实 施例中, 仅以待处理圈像为 RGB H像为例进行滅明。 下结合图 2主要 «迷上 述步骤 103和歩骤 105的具体实现 ^程
在通过歩骤 102获取高 滤波图像之后, MBh 根据所逑高斯滤波图像的 各个像素点的灰度, 按照灰度从小到大 順序确定各个灰度所对 W像素点个 数。
对高斯滤波圍像的各个灰度对雍翁像素点个数翁统计相当于对该高辦滤波 圈像逾行直方图统计, 據厘灰度从小到大的順序, 获取各个灰度对 像素点 个数, 该直方图的撗坐标轴为灰度, 纵 标轴为灰度对産的像素点个数
1032: 根据所述高斯滤波團像的像素点个数和预 i 比例, 确定第一数值。 其中, 所違高斯滤波图像翁像素点个数和待处理图像的像素点个数需保持 一致, ώ于在后续歩骤中邇过高插滤波图像获取的灰度变化率用来处理待处理 图像得来, 所以需要使高斯滅波闻像的 像^点个数与待处理團像 W总像素点 个数相同。 若待处理图像 W大小是 Μ行 Ν列, 则待处理图像的像素点个数是 ΜχΝ, 高斯滤波圏象的像素点个数 应为 MxN, 该对 H像总像素点个数《处理 可以在歩骤 202的滤波过程中实现。
将高斯滤波圍像的像素点个数和预資比倒进行乘积运算, 获取第一数值, 第一数值可用于获取 T述的第- 灰度, 迸丽可以区分圈像中 WW影区域和卨光 区域 δ 该预 比俩可以迸行根据圈像的实 1¾情«进行 ¾置, 若 S像中的明影区 域较大, 高光区域较小, if以将镇预设比倒 ¾«的较人, 如 80%, 若图像中 的阴影区域较小, 高光区域较大, 可以将 預谁比倒设置曲较小》 如 20¾。
1033: 根据所逑第一数值和确定的所逾各个灰度所对産的像素点个数, 获 取第一灰度。
其中, 厨述第一灰度用于区分 H像的 影区域和高光区域。 为了较好地调 节 ffl像的亮度, 将小于第一灰度 像素点诈为囊影区域, 将大于第一灰度的像 素点作为高光区域, 以便于对两个 E域分劃 行亮度的调节。 在确定第 - ·· -数 ί
后, 按照灰度从小到大的順序对高斯滤波图像的各个灰度 if对应的像素点个数 进行的累加, 直到前 次的累加僮小于第一数值, 丽后一次 累加值大亍等于 第一数值时, 停 It累加, 将后一次的累加的像素点 个数在直方圈上对应的灰 度作为第一灰度„
上逮歩骤 1031 1033是植据 逮高斯漶波图像的像素点个数、各个像素点 灰度以及预设比例, 获取用于区分 H像的 W影区域和高光区域的第一灰度的过 程的 ·····个示例《
1034: 根据所逮高斯滤波圏像的各个像素点的灰度和所述第一灰度, 调节 所述高斯滤波團像的灰度 β
在一个示例中, 当高斯滤波圍偉的第一像素点的灰度小于第一灰度时, 则 廣途第一像素点组成的区域为明影区域,
馬 S
像素点脑灰度, 其中, p 是高斯減 图像 第一像素点的位置 Φ标, 2( )是 «节后第一像素点的灰度, 是调节前的第一像素点的灰度 是预设參数,
5>0, HI是第一灰度;
其中, 第一像素点是指高》滅波图像中的任一个像素点, 上述公式 ÷¾用 的 ||表示第- 像素点的位. W坐标, 若高斯滤波图像的大小为 M ^ N列, 则 p的 取值 ¾B为『1, MxN]0对于二维图像的位置坐标, 可以用两个分量数据来表示, 也可以用 -个量来表 , 如对于大小为 M ^ N列的图像, 第 行第/列的像素 点的位置 标可以表示为 /], 也可以表示为 Mxi'+j', 这两种表示方式可以相 互转化, 在本发嚷实施倒中仅以一个数握表示像素点的位置坐标为例 a行 明 , 在实 应用中可以采 一个数据或者两个分量数据, 本发嚷实施侧对此不作具 体限定
采用上述公式遍历高猜滤波 H像中灰度小于第一灰度 像素点, 实現 T对 圖影区域 调节。 通过利用上途非錢 tt变换公式对图像中 ttPJ影区域中的像素
点的灰度进行非线性的调节后, 使得阴影区域的亮度增大, 局部对比度增强, 细节信息突出, 便于人眼的观察。
当高斯滤波图像的第一像素点的灰度等于第一灰度时, 不对所述第一像素 点的灰度进行调节。
因为, 当高斯滤波图像中的第一像素点等于第一灰度时, 说明第一像素点 的灰度适合于人眼的观察, 亮度适中, 所以不对这些像素点的灰度进行调节。 即-
L2(P) = (p) (4)
其中, p 是高斯滤波图像中的第一像素点的位置坐标, ^(p)是调节后第一 像素点的灰度, 是调节前的第一像素点的灰度。
当高斯滤波图像的第一像素点的灰度大于第一灰度时, 则所述第一像素点
(1-Lx{p)) x (\ + -)
组成的区域为高光区域, 应用 £2(^ = 1 -~ Γ_ΓΓπ ·调节所述第一像素点的灰
J
1— h
度, 其中, p 是高斯滤波图像中的第一像素点的位置坐标, ^p)是调节后第一 像素点的灰度, 是调节前的第一像素点的灰度, A是预设参数, A>0, 是 第一灰度。
采用上述公式遍历高斯滤波图像中灰度大于第一灰度的像素点, 实现了对 高光区域的调节。 通过上述非线性变换公式对图像中的高光区域中的像素点的 灰度进行非线性的调节后, 使得阴影区域的亮度减小, 局部对比度增强, 细节 信息突出, 便于人眼的观察。
步骤 1031-1034是上述步骤 103 的根据所述高斯滤波图像的灰度和预设比 例, 调节所述高斯滤波图像的灰度的具体过程的一种示例。
以下描述步骤 105的具体过程。
在获取调节前后每个像素点的灰度变化率之后, 1051 : 将所述每个像素点 的灰度变化率和所述待处理图像中的对应的像素点的灰度相乘, 得到第二灰度。
10
替换页 (细则第 26条)
tlj于待处理图像为 RGB网像, 因此需要对待处理图像的 R、 G、 B三个通 道勺灰度分别进行处理, 将每个像素点的灰度变化率和各个通道 W对应的像素 点翁灰度 ¾行相乘, 可以获取 ¾每个通通每个像素点 第二灰度 '
1052: ft—像素点翁第二灰度大于第二预 僮时, 将所述 «—像素点的 灰度调节为第二预设值。
在一个示例 Ψ, 第二预设值的设置需要极据 a像采用 数值类靈《行相魔 翁设置。 使用无符号 8位整靈(uint8 )即 0 255表示 »图像灰度时, 第二預 ¾ 值可以设置为 255, 一像素点的第二灰度大于 255时, 将 255作为该任一像 素点的灰度。 使用取精度(double) 即 (M表示的 W像灰度时, 第二预设值 ¾! 以 置为 1 , ¾ 一像素点曲第二灰度大于 1时,将 1作为该 ft—像素点的灰度 由于灰度变化率的范围较大, 当灰度变化率大于 1 时, 获取 第二灰度可能超 出图像的灰度显示菌围。 在实 应用中, 需要根据具体 数值类 ¾¾置第二预 谁值, 所以, 本发 实施倒对第二预设值的大小不作具律限定。
通过第二预设值对圈像灰度 Wi周整, 使得 像翁灰度菌围保持在圈像可显 示的 ¾围之内。
1053 将调节后的團像获取为经 ¾处理后 團像。
歩骤 1051 -1053是上述歩骤 105的根据 述每个像素点 灰度变化率对待处 理 a像 a行处理, 获取经过处理后翁图像的具体过程 一种示例。
在本发嚷翁另一实施倒中, 以仅通过将所述每个像素点 灰度变化率 和所述待处理图像中的对应的像素点的灰度相乘来对所遣待处¾圈像 a行处 理。 待处理图像经过一系列的处理, PJ影区域和高光区域得 ¾了调节, 调节后 I»H像 s合于人眼的观察 及細节信息的获取。
本发嚷实施側通过获取待处理 ¾像》 根据所述待处理圈像 各个通遒 ό 灰 度获取单通道亮度图像; 对所述单通道亮度图像《行高斯滤波, 获取高斯滤波 图像; 根据所述高斯滤波图像的灰度和预 t倒, i周节所建高斯滤波图像的灰 度; 比较调节后的高斯滅波圈像和谓节前 高斯滤波图像, 获取谓节前 每个
像素点的灰度变化率; 通过将所述每个像素点的灰度变化率和所逮待处理 B像 中的对魔的像 · 点的灰魔相乘来对所述待处理圍像逬行处理; 输出经过处囊后 的图像, 来对图像各个通道的灰度的处理, 本发 BJ实施倒提供的技术方案充分 地利用了整个颜色通道的表迖能力, 通过对旱遣道亮度图像逬行高斯滤波, 保 证了圏像整体的连续性, 通过根据预设比例调节高斯滤波圈像, 使高斯滤波 ffl 像中闘影和 '光区城趁于中性, 更适合于入眼的观察。
图 3 是根揚本发明实施倒中提供的一 tffl像亮度调节装置的结祷框圏, 参 圈 3, 该装置包括:
第一获取模块 301,用于获取待处理圍像, 根掘所逮待处理 ffl像 各个通道 的灰度获取单通道亮度图像。
具体地, 第一获取模块 301 用于获取待处理图像, 将每个像素点的各个亮 産通道的灰度进行比较, 获取所建每个像素点灰度的最大值, 将所述最大值作 为所逮每个像素点的灰度, 得到单通道亮度图像。
第二获取模块 302, 用于对 建单通道亮度图像遣行高斯滤波, 获取卨斯滤 波图像。
谓节模块 303, 用于根据所 高斯滤波图像的灰度和预设比例, 调节 述高 斯滤波 ,像的灰度。
在一个 例中, 如图 4所示, 述调节模块 303包括:
第一灰度获取旱元 3031, 用予根揚所逾高斯滤波 ffl像的像素点个数、 各个 像素点的灰度以及预 ¾比例, 获取用于区分图像的明影区域和高光区域的第一 灰度。 .
具体地, 述第一灰度获取攀元用于根据所述高斯滤波圍像 各个像素点 的灰度, 接照灰度从小到大 »順序确定各个灰度 对应 W像素点个数; 根据所 述高斯滤波 ^像的像素点个数和预设比例, 确定第一数值; 根摇所通第一数值 和确定的所述各个灰度所对廣 W像素点个数, 获取第一灰度。
第一侧节单元 3032, 用丁根据所述高斯滤波图像 各个像素点的灰度和
述第一灰度, i肩节所建高斯滤滅 B像的灰度。
具#地, 所建第一调节单元用于^高斯滤 ¾图像的第一像素点的灰度小于 第一灰度时,则所述第一像素点组成的区域为 κ影区域,应用 £2(,) () )(i
腿 S 调节 述第一像素点的灰度; 当高斯滤波 H像的第一像素点的灰度等于第一灰 度时, 不 所述第一像素点的灰度 a行谓节; ¾高斯滤 nil像 »第一像素点的 灰度大于第一 , 则所逮第一像素点纖成的 E域为高光区域, 应用 其中, 是高康滤波圈象
中的第一 ·像素点 W位置坐标, £2( )是调节后第一像素点的灰度, £ (| )是《节前 »第一像素点的灰度, s是预设参数, >0. 應是第 - ··爐, 是预设参数, Λ>0。
第三获取模块 304,用于比较¾节后的高斯滤波 ffl像和慣节前的高斯滤波圈 像, 获取调节前后每个像素点的灰度变化率。
第四获取模块 305,用于棍据所建每个像素点的灰度变化率对所述待处理图 像进行处理, 获取δΜ:处理后的圈像。
在一个示侧中, 如图 5所示, 第四获取模块 305包括- 第二灰度获取单元 3051 , 用于将所述每个像素点的灰度变化率和所逑待处 理 Η像中 对应的像素点 Κ灰度相乘, 得到第二灰度;
第二调节单元 3052, 用于¾任一像素点的第二灰度大于第二預设僮时, 将 所建任一像素点 W灰度侧节为第二预设值;
经过处理后 S像获取单元 3053, 用亍将调节后的图像获取为经过处理后 的圏像。
输出模块 306, 用于输出经过处理后的圈偉。
本发明实施例的装置遣 ¾±第一获取模块 301 获取待处理图像, 根据所逮待 处理圈像 W各个 ¾道的灰度获取单通道 ¾度图像; 第二获取模块 302对所述单
通道亮度图像 ¾行高斯滤波, 获取高斯滤波圍像; 调节模块 303棍据所述高斯 滤波 ffl像的灰度靡预设比俩, 《节所 高 滤波图像 W灰度; 第三获取模块 304 比较囊节后 高 »«波圈像和 i厕节前 高斯滤波 ffl像, 获取调节前后每个像素 点的灰度变化率; 第四获取模块 305 根据所述每个像素点的灰度变化率对所逮 待处理圈德进行处理, 获取经过处理后 图像; 输出模块 306输出经 a处理后 的围像来对 a像各个通遊的灰度 处理, 本发嚷实施例提供的 s术方案充分地 利用了整个颜色通道 表达能力, 通过对单通道亮度園像迸行高斯滤波, 保证 了園像整 的连续性》 a过棍据预 比侧 i周节高斯滤波圈像, 使高斯滤波图像 中阴影和高光区域趋于中性, 更适合于人眼 »观察。
需要说明翁是: 上述实施侧提供的圏像亮度调节装置 处理圈像的 κ影和 高 区域时, 仅以 h 各功能模块的划分 行举倒说明, 实 应用中, ¥以根 据 要 ffi将上述功能分配由不同 w功能模块完成, 即将 备 内部结枸划分成 不同的功籠模块, 存储在计算设备翁存 «装置中, 被该计算设备中 I»至少一个 处理器执 if, t¾完歲以 h描建的全部或者部分功能。 另外, ..... t逑实施俩提供 图像亮度调节装 s与图像亮度调 w方法实施例属予同一构愿, 其具律实现过程 见方法实施例, 这里不再赞建。
本領域普通技术人员 以理解实現上逑实施侧的全部或部分歩骤¥以通过 硬件 *完成, 也可 K遣过程序来措令相关的硬件完成, 所建的程序可以存储亍 一种计算机可读存储介质中, 上¾提到 存储介质可 是只读存储器, 磁盘或 光盘等。
以― t所述仅为本发明的较德实施例, 并不用 K限制本发明, 凡在本发明的 精神和原则之内, 作的任何修改、 等同替换、 改进等, 均 包含在本发明的 保护舊围之内 β
Claims
1、 一种 ffl像亮度调 ϊ方法, 其特征在于, 所述方法包括:
获取待处理图像, 根据 逮待处理圍像 各个通道 W灰度获取单通道亮度 ffl像;
对所達单通道亮度圈像进行高斯滤波, 获取高 if滤波 s像;
權据所逮高斯滤波 a像的灰度和预谁比倒, 调节所述高斯滤波图像 灰度; 比较调节后的高斯滤波 ffl像和调节前 w高斯滤波園像, 获取调节前后每个 像素点的灰度变化率;
根据所述每个像素点的灰度变化率对所途待处囊團像 a行处理, 获取经过 处理后 MB像;
输出经过处理盾的圈像
2、 根据极 要求 1所述 ¾方法, 其特征在于, 获取待处理图像, 根据所逮 待处理图像 w各个邇道的灰度获取单通道亮庋闕像, 包括- 获取待处理 a像, 将每个像素点的各个亮度通道曲灰度進行比 ¾;
获取所逮每个像素点灰魔》最大僮, 将所述 大值作为所建每个像素点的 灰度, 得到单通道亮度 a像
3、 根摇权利要求 1所途》方法, 其特征在于, 根据所述卨斯滤波圈像的灰 度和预设比例, 调节所述高斯滤波圈像的灰度, 包括:
根据所 高 »滤波圍像的像素点个数、 各个像素点的灰度以及预设 tt例, 获取用于区分圏像的 K影区域和高光区域的第一灰度;
根据所逮高斯滅波圈像的各个像素点 灰度和廣途第一灰度, 谓节所述高 斯滤波圍像 tt灰度。
4、 根据权利要求 3所逑的方法, 其特征在于, 根据所述高斯滤被圍像的像 素点个数、 各个像素点 灰度以及预 i 比例, 获取 于区分圈像翁園影区域和 高光区域 第一灰度, 包括:
根据所逑高斯滤波图像的各个像素点的灰度, 按照灰度从小到大的順序确 定各个灰度所对应 W像素点个数; '
根据所述高斯滤波 H像 像素点个数和预 i 比例, 确定第一数值; 根«所述第 数僮和确定的所述各个灰度所对应的像素点个数, 获取第一 灰度。
5、 根攝权《要求 4所逑的方法, 其特征在于, 根据厨述第一数值和确定 所述各个灰度所对 的像素点个数, 获取第一灰度, 包 Ί
对确定的所 ¾各个灰度所对应的像素点个数进行累加
当前一次的累加值小 f所建第一数值、 后- - -次的累加值大于等子所述第一 数值吋, 将所述后一次的累加 W像素点个数对应的灰度作为第一灰度。
6、 根据权利要求 3所述的方法, 其特征在于, 根据高辦滤波圍像的各个像 素点的灰度, 调节所建高斯滤波图像 灰度, 包揺:
当高斯滤波 H像 第- -像素点的灰度小于第- - -灰度时, 则 述第一像素点 组成的区域为闘影区域,
高斯滤波图像的第- 象素点 灰度等于第一灰度 Bt, 不对所逮第一像素 点 W灰度进行调节;
当高斯滤波 H像翁第一像素点的灰度大于第- - -灰度时, 員續述第一像索点 组戚 区域为高光区域,
7、 棍摇权利要求 3所述的方法, 其特征在丁 ·, 根据所達每个像素点的灰度 变化率对厨逑待处理 H像逾行处现, 获取经过处理后的图像, 包括- 将所述毎个像素点的灰度变化率和所逮待处理图像 Ψ的对应的像素点的灰 度相乘, 得到第二灰度;
当任一像素点 W第二灰度大于第二预 ¾僮时, 将所逮任一像素点的灰度调 节为第二预设僮;
将调节后的 1 像获取为经过处理后的图像。
8、 一种图像亮度调节装肯, 其 征在于, 廣 装置包括:
第一获取模块》 用于获取待处理图像, 棍据 途待处理 ffl像的各个通道的 灰度获取单通道亮度圏像;
第二获取模块, 用于对所述单通道亮度 ffl像逮行高斯滤波, 获取高斯滤波 图像;
调节摸块, 用亍根攝所 ft斯減波圈像的灰度和预设比例, 调节廣述高斯 滤波图像的灰度;
第三获取模块, 用于比较调节后的高斯滤波圏像和调节前的高斯滤波 S像, 获取谓节前后每个像素点的灰皮变化率:
第四获取模块, 用于根据所逑每个像素点的灰度变化率对所建待处理圏像 纖行处理, 获取经过处理后的圈像
输出模块, 用于输出经过处理后的圈像
9、 根据 ft 要求 8所述的装覽, 其特 fr:在于, 所述第一获取模块用于获取 待处雇图像, 将每个像索点 各个亮度通道的灰度 a行比较, 获取廣述每个像 索点灰度的 it大值, 将所述最大值作为所述每个像素点 灰度, 得到单通道亮
10、 很据权利要求 8所¾的装置, 其特怔在 f , 所述调节模块包¾:
第一灰度获取单元, 用于根 «所逮高簾滤液 H像的像素点个数、 各个像素 点的灰度以及预设比侧, 获取用亍区分 像» 影区域和高光区域的第一灰度; 第一 i肩节单元, 用于根据所述? 斯滤纖 B像曲各个像素点的灰度和所 第 一灰度, i厕节所述 ft斯滤波图像的灰度。
11、 棍据权利要求 10 述的装置, 其特征在于, 所逑第一灰度获取单元用 于根据所述高斯滤波图像 W各个像素点 «I灰度, 拔皿灰度从小到大的順序确定 各个灰度所对应的像素点个数; 根«所述高斯滤波图像 像素点个数和预设比 例, 确定第一数值; 根据 逮第一数值和确定 達各个灰度 If对! 像素点 个数, 获取第一灰度
12、 根振权剩要求 11所逮的装置, 其特征在于, 述第一灰度获取单元用 子根据所述高斯滤波圏像 W各个像素点的灰度, 抜皿灰度从小到大 順序确定 各个灰度所对应的像:秦点个数; 极攝 述高麵滤波图像的像素点个数和预设比 例, 确定第一数值; 对确定 W所述各个灰度 对应的像素点个数进行累加; 当 前一次的累加值小于所逑第-数值、 后一次 W累加值大亍等于所逑第一数值时, 将廣述后一次翁累加 W像素点个数对康的灰度作为第一灰度《
謂 S
滤波 Η像 第一像素点的灰度等于第一灰度时, 不对所述第一像 点的灰度纖 行调节; 当高 滤波图像的第一像素点的灰度大于第一灰度时, 则所述第一像 素点组成的区域为高光区域,
«灰度, 其中, p 是高斯滤波图像 Ψ«Ι第一像素点的位置 标, £2( )是调' 1S»7S 第一像素点 灰度, 是调节前翁第一像素点的灰度, s是预设参数, 5>0, m是第一灰度, 是预设参数, A>0o
14、 根据极利耍求 10所述的装置, 其特征在于, 所述第圆获取模块包括: 第二灰度获取单元, 用亍将所逮每个像素点的灰度变化率和所述待处理 ffl 像中 对应 W像素点 灰度相乘, 得到第二灰産;
第二调节单元, 用千当任一像素点的第二灰度大于第二预 億时, 将 f述 任一像素点的灰度调节为第二预设值;
经过处理后的图像获取单元, 用子将调节后的图像获取为经过处理后的图 像。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US14/806,142 US9824430B2 (en) | 2013-01-23 | 2015-07-22 | Method and apparatus for adjusting image brightness |
| US15/786,049 US10074164B2 (en) | 2013-01-23 | 2017-10-17 | Method and apparatus for adjusting image brightness |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310025149.0A CN103942755B (zh) | 2013-01-23 | 2013-01-23 | 图像亮度调节方法和装置 |
| CN201310025149.0 | 2013-01-23 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US14/806,142 Continuation US9824430B2 (en) | 2013-01-23 | 2015-07-22 | Method and apparatus for adjusting image brightness |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2014114126A1 true WO2014114126A1 (zh) | 2014-07-31 |
Family
ID=51190407
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2013/087193 Ceased WO2014114126A1 (zh) | 2013-01-23 | 2013-11-15 | 图像亮度调节方法和装置 |
Country Status (3)
| Country | Link |
|---|---|
| US (2) | US9824430B2 (zh) |
| CN (1) | CN103942755B (zh) |
| WO (1) | WO2014114126A1 (zh) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966271A (zh) * | 2015-05-28 | 2015-10-07 | 电子科技大学 | 基于生物视觉感受野机制的图像去噪方法 |
| CN111950558A (zh) * | 2020-08-26 | 2020-11-17 | 上海申瑞继保电气有限公司 | 高压油浸式变压器油液液位图像识别方法 |
| CN113344809A (zh) * | 2021-05-27 | 2021-09-03 | 同济大学 | 一种超声图像增强方法、系统和设备 |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104992423B (zh) * | 2015-07-14 | 2019-05-03 | 厦门美图之家科技有限公司 | 一种基于梯度的图像锐化处理方法、系统及拍摄终端 |
| CN105227805A (zh) * | 2015-09-28 | 2016-01-06 | 广东欧珀移动通信有限公司 | 一种图像处理方法及移动终端 |
| CN106101568B (zh) * | 2016-08-16 | 2021-06-01 | 北京中星微电子有限公司 | 一种基于智能分析的强光抑制方法及装置 |
| CN107730460B (zh) * | 2017-09-26 | 2020-02-14 | 维沃移动通信有限公司 | 一种图像处理方法及移动终端 |
| CN111046887A (zh) * | 2018-10-15 | 2020-04-21 | 华北电力大学(保定) | 一种带噪图像特征提取的方法 |
| CN109584182A (zh) * | 2018-12-03 | 2019-04-05 | 董育理 | 一种图像处理方法和系统 |
| CN110490832B (zh) * | 2019-08-23 | 2023-05-05 | 哈尔滨工业大学 | 一种基于正则化深度图像先验方法的磁共振图像重建方法 |
| CN111027612B (zh) * | 2019-12-04 | 2024-01-30 | 国网天津市电力公司电力科学研究院 | 基于加权熵fcm的能源计量数据特征约简方法及装置 |
| CN111311496B (zh) * | 2020-03-25 | 2023-03-21 | 厦门汉印电子技术有限公司 | 图像分辨率快速调整方法和装置 |
| CN112365413B (zh) * | 2020-10-30 | 2024-07-26 | 湖北锐世数字医学影像科技有限公司 | 图像处理方法、装置、设备、系统及计算机可读存储介质 |
| CN113052656B (zh) * | 2021-03-29 | 2022-02-22 | 蔡文华 | 一种基于大数据的电商平台管理系统 |
| US12387669B2 (en) * | 2022-12-16 | 2025-08-12 | Apple Inc. | Display panel brightness control based on gaussian edges |
| CN116129806B (zh) * | 2022-12-28 | 2025-04-15 | 厦门天马显示科技有限公司 | 显示面板的控制方法及控制装置、存储介质及电子设备 |
| CN116740809B (zh) * | 2023-06-05 | 2024-03-29 | 嘉兴米兰映像家具有限公司 | 一种基于用户姿态的智能沙发控制方法 |
| CN116402816B (zh) * | 2023-06-08 | 2023-08-15 | 中国人民解放军海军青岛特勤疗养中心 | 一种体检ct影像数据的管理方法及系统 |
| CN117788570B (zh) * | 2024-02-26 | 2024-05-07 | 山东济矿鲁能煤电股份有限公司阳城煤矿 | 一种基于机器视觉的斗轮机定位方法及系统 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070046010A (ko) * | 2005-10-28 | 2007-05-02 | 삼성전자주식회사 | 디지털 이미지 개선방법 |
| CN101437114A (zh) * | 2008-12-12 | 2009-05-20 | 深圳市斯尔顿科技有限公司 | 一种图像亮度的调节方法及系统 |
| CN102411775A (zh) * | 2011-08-10 | 2012-04-11 | Tcl集团股份有限公司 | 一种校正文本图像亮度的方法和系统 |
| CN102750535A (zh) * | 2012-04-01 | 2012-10-24 | 北京京东世纪贸易有限公司 | 自动提取图像前景的方法和系统 |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8014034B2 (en) * | 2005-04-13 | 2011-09-06 | Acd Systems International Inc. | Image contrast enhancement |
| US8164628B2 (en) * | 2006-01-04 | 2012-04-24 | Mobileye Technologies Ltd. | Estimating distance to an object using a sequence of images recorded by a monocular camera |
| US8023733B2 (en) * | 2006-06-08 | 2011-09-20 | Panasonic Corporation | Image processing device, image processing method, image processing program, and integrated circuit |
| KR100791387B1 (ko) * | 2006-10-25 | 2008-01-07 | 삼성전자주식회사 | 영상 처리 방법 및 장치 |
| JP4577583B2 (ja) * | 2007-12-04 | 2010-11-10 | ソニー株式会社 | 画像処理装置および画像処理方法、プログラム、並びに、記録媒体 |
| CN103761134A (zh) * | 2014-01-28 | 2014-04-30 | 华为技术有限公司 | 一种处理图片的方法及电子设备 |
| US9712720B2 (en) * | 2014-06-02 | 2017-07-18 | Intel Corporation | Image refocusing for camera arrays |
-
2013
- 2013-01-23 CN CN201310025149.0A patent/CN103942755B/zh active Active
- 2013-11-15 WO PCT/CN2013/087193 patent/WO2014114126A1/zh not_active Ceased
-
2015
- 2015-07-22 US US14/806,142 patent/US9824430B2/en active Active
-
2017
- 2017-10-17 US US15/786,049 patent/US10074164B2/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070046010A (ko) * | 2005-10-28 | 2007-05-02 | 삼성전자주식회사 | 디지털 이미지 개선방법 |
| CN101437114A (zh) * | 2008-12-12 | 2009-05-20 | 深圳市斯尔顿科技有限公司 | 一种图像亮度的调节方法及系统 |
| CN102411775A (zh) * | 2011-08-10 | 2012-04-11 | Tcl集团股份有限公司 | 一种校正文本图像亮度的方法和系统 |
| CN102750535A (zh) * | 2012-04-01 | 2012-10-24 | 北京京东世纪贸易有限公司 | 自动提取图像前景的方法和系统 |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966271A (zh) * | 2015-05-28 | 2015-10-07 | 电子科技大学 | 基于生物视觉感受野机制的图像去噪方法 |
| CN111950558A (zh) * | 2020-08-26 | 2020-11-17 | 上海申瑞继保电气有限公司 | 高压油浸式变压器油液液位图像识别方法 |
| CN111950558B (zh) * | 2020-08-26 | 2024-02-09 | 上海申瑞继保电气有限公司 | 高压油浸式变压器油液液位图像识别方法 |
| CN113344809A (zh) * | 2021-05-27 | 2021-09-03 | 同济大学 | 一种超声图像增强方法、系统和设备 |
Also Published As
| Publication number | Publication date |
|---|---|
| US20180053290A1 (en) | 2018-02-22 |
| US9824430B2 (en) | 2017-11-21 |
| HK1200580A1 (zh) | 2015-08-07 |
| US10074164B2 (en) | 2018-09-11 |
| CN103942755A (zh) | 2014-07-23 |
| US20150324961A1 (en) | 2015-11-12 |
| CN103942755B (zh) | 2017-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2014114126A1 (zh) | 图像亮度调节方法和装置 | |
| US9691139B2 (en) | Adaptive contrast in image processing and display | |
| US8941762B2 (en) | Image processing apparatus and image pickup apparatus using the same | |
| US9824424B2 (en) | Image amplifying method, image amplifying device, and display apparatus | |
| US9858666B2 (en) | Medical skin examination device and method for processing and enhancing an image of a skin lesion | |
| US7672528B2 (en) | Method of processing an image to form an image pyramid | |
| CN107993189B (zh) | 一种基于局部分块的图像色调动态调节方法和装置 | |
| US20160140696A1 (en) | Image processing device, image processing method, and image processing program | |
| CN106504278A (zh) | 高动态范围色调映射 | |
| EP3864616B1 (en) | Systems and methods for image processing | |
| CN103873740B (zh) | 图像处理装置及信息处理方法 | |
| JP2024037722A (ja) | コンテンツに基づく画像処理 | |
| US20120133801A1 (en) | Information processing apparatus and method | |
| JP2016046706A (ja) | 画像処理装置、画像処理方法、記憶媒体及びプログラム | |
| US10868969B1 (en) | Method and apparatus for accelerated tonemapping and display | |
| CN114998122A (zh) | 一种低照度图像增强方法 | |
| US9202264B2 (en) | Image processing apparatus and image processing method | |
| US9053552B2 (en) | Image processing apparatus, image processing method and non-transitory computer readable medium | |
| JP5682443B2 (ja) | 画像処理装置、画像処理方法及び画像処理プログラム | |
| CN104504659B (zh) | 一种基于提升小波变换的快速iso去噪方法及系统 | |
| US20140169694A1 (en) | Image processing apparatus and image processing method | |
| CN110689486A (zh) | 图像的处理方法、装置、设备及计算机可存储介质 | |
| CN108810320A (zh) | 图像质量提升方法和装置 | |
| RU2477007C1 (ru) | Система коррекции темных, светлых и средних тонов на цифровых изображениях | |
| EP4610922A1 (en) | Training method, training apparatus, image processing method, image processing apparatus, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13872296 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC, EPO FORM 1205N DATED 29.09.2015 |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 13872296 Country of ref document: EP Kind code of ref document: A1 |





