WO2020093600A1 - 基于图像多重曝光融合的高动态显示方法 - Google Patents

基于图像多重曝光融合的高动态显示方法 Download PDF

Info

Publication number
WO2020093600A1
WO2020093600A1 PCT/CN2019/072435 CN2019072435W WO2020093600A1 WO 2020093600 A1 WO2020093600 A1 WO 2020093600A1 CN 2019072435 W CN2019072435 W CN 2019072435W WO 2020093600 A1 WO2020093600 A1 WO 2020093600A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
fusion
exposure
detail
layer
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
Application number
PCT/CN2019/072435
Other languages
English (en)
French (fr)
Inventor
史超超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
TCL China Star Optoelectronics Technology Co Ltd
Original Assignee
Shenzhen China Star Optoelectronics Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen China Star Optoelectronics Technology Co Ltd filed Critical Shenzhen China Star Optoelectronics Technology Co Ltd
Publication of WO2020093600A1 publication Critical patent/WO2020093600A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • 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
    • G06T5/90—Dynamic range modification of images or parts thereof
    • 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
    • G06T5/94—Dynamic range modification of images or parts thereof based on local image properties, e.g. for local contrast enhancement
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20221—Image fusion; Image merging

Definitions

  • the invention relates to a high dynamic display method based on image multiple exposure fusion, which enhances the image in different regions, respectively enhances the human eye interest area of different exposure images, and retains more details in the human eye interest area To enhance the overall visual effect of the image.
  • Digital cameras have the advantage of taking pictures quickly without the need to develop negatives, so they have replaced traditional cameras for many years.
  • images captured by digital cameras are prone to overexposure in high light and dark in low light. Therefore, many people have developed a multi-exposure fusion method that performs digital post-processing on camera images to obtain highly dynamic images.
  • the prior art multi-exposure fusion method generates multiple exposure images by establishing a suitable exposure function on one image.
  • the weight of each exposure image is calculated to obtain the average value of the images as the central value.
  • each different exposure image tends to focus on different points, for example, the darker exposure image tends to focus on the brightest area (such as the sky). Conversely, the brightest exposed image needs to be enhanced in darker area details. Therefore, the average value of the prior art can not obtain a better image display effect, and the resulting image is often whitish or blurred.
  • the invention provides a high dynamic display method based on image multiple exposure fusion to solve the problem that the image in the prior art is whitish or blurred after processing.
  • the main objective of the present invention is to provide a multi-exposure fusion high-dynamic display method, including:
  • the multi-exposure image generation step includes generating multiple exposure images from the original image using a suitable S-shaped function
  • the human eye interest area information extraction step includes extracting multiple human eye interest areas in each of the exposure images through an image saliency model
  • the step of calculating the weight of the human eye interest area includes calculating the weight value of each of the human eye interest areas in each of the exposure images separately;
  • the extraction step of the base layer and the detail layer includes extracting the image base layer and the image detail layer from each of the exposed images;
  • the step of fusing images includes fusing all image base layers to generate a fusion base layer, fusing all image detail layers to generate a fusion detail layer, and finally fusing the fusion base layer and the fusion detail layer to generate a fusion image.
  • the method further includes: a computer providing step, including providing a computer; an original image input step, including inputting the original image to the computer; the multi-exposure image generating step, the person The eye interest region information extraction step, the human eye interest region weight calculation step, the base layer and detail layer extraction step, and the fused image step are performed by the computer operation.
  • the multi-exposure fusion high dynamic display method further includes an output step, and the output step includes outputting the fusion image to an external electronic device through the computer.
  • the exposure values of the multiple exposure images are different.
  • the plurality of exposed images are grayscale images.
  • the base layer and detail layer extraction step uses a principal component analysis method to extract the image base layer and the image detail layer from each of the exposed images.
  • the step of fusing the image includes the step of fusing the image base layer, which includes performing, by the computer operation, according to the weight value of each region of interest of the human eye in each of the exposed images
  • the multiple image base layers are weighted and fused to generate the fused base layer.
  • the step of fusing images further includes a step of fusing image detail layers, including fusing a plurality of image detail layers through the computer operation to generate the fusion detail layer, wherein
  • the image detail layer fusion step includes first generating a plurality of detail enhancement coefficients, and then fusing the plurality of image detail layers through a detail layer fusion calculation formula.
  • the step of fusing images further includes a step of generating a fusing image, including fusing the fusing base layer and the fusing detail layer through the computer operation to generate the fusing image.
  • the detail enhancement coefficient 1 + sqrt (std (L) / 255-std (B) / 255); where L represents the gray value of the exposed image of the original image and B represents the basis The gray value of the layer, sqrt stands for square root and Std stands for standard deviation.
  • the fusion base layer and the fusion detail layer are fused into the fused image through linear superposition.
  • Another object of the present invention is to provide a multi-exposure fusion high dynamic display method, including:
  • the multi-exposure image generation step includes generating multiple exposure images from the original image using a suitable S-shaped function
  • the human eye interest area information extraction step includes extracting multiple human eye interest areas in each of the exposure images through an image saliency model
  • the step of calculating the weight of the human eye interest area includes calculating the weight value of each of the human eye interest areas in each of the exposure images separately;
  • the extraction step of the base layer and the detail layer includes extracting the image base layer and the image detail layer from each of the exposed images;
  • the step of fusing images includes fusing all image base layers to generate a fusion base layer, fusing all image detail layers to generate a fusion detail layer, and finally fusing the fusion base layer and the fusion detail layer to generate a fusion image;
  • the exposure values of the multiple exposure images are different;
  • the multiple exposure images are grayscale images
  • the step of extracting the base layer and the detail layer is to extract the image base layer and the image detail layer from each of the exposed images using a principal component analysis method.
  • the method further includes: a computer providing step, including providing a computer; an original image input step, including inputting the original image to the computer; and the multi-exposure image generating step, the The human eye region of interest information extraction step, the human eye region of interest weight calculation step, the base layer and detail layer extraction step, and the fusion image step are performed by the computer operation.
  • the multi-exposure fusion high dynamic display method further includes an output step, and the output step includes outputting the fusion image to an external electronic device through the computer.
  • the step of fusing the image includes the step of fusing the image base layer, which includes performing, by the computer operation, according to the weight value of each region of interest of the human eye in each of the exposed images
  • the multiple image base layers are weighted and fused to generate the fused base layer.
  • the step of fusing images includes the step of fusing image detail layers, including fusing multiple image detail layers through the computer operation to generate the fusion detail layer, wherein the
  • the image detail layer fusion step includes first generating a plurality of detail enhancement coefficients, and then fusing the plurality of image detail layers through a detail layer fusion calculation formula.
  • the fusion image step includes: a fusion image generation step, which includes fusing the fusion base layer and the fusion detail layer through the computer operation to generate the fusion image.
  • the detail enhancement coefficient 1 + sqrt (std (L) / 255-std (B) / 255); where L represents the gray value of the exposed image of the original image and B represents the basis The gray value of the layer, sqrt stands for square root and Std stands for standard deviation.
  • the high dynamic display method of image multiple exposure fusion of the present invention generates multiple exposure images with different exposure values from the original image, and then extracts the image base layer and image details for each of the exposure images Layer, then fuse the multiple image base layers with a specific weight value, and fuse the multiple image detail layers with specific detail enhancement details, and finally fuse the fusion base layer and the fusion detail layer a second time to obtain a fusion image .
  • the fusion image obtained by the method of the present invention has been enhanced in detail for the region of interest of the human eye, so compared with the prior art image enhancement method optimized according to the average value of the image, the present invention has a better display effect.
  • FIG. 1 is a block diagram of a computer architecture to which the high dynamic display method of image multiple exposure fusion of the present invention is applicable.
  • FIG. 2 is a block diagram of the image processing flow of the high dynamic display method of image multiple exposure fusion of the present invention.
  • FIG. 3 is a flowchart of the steps of the high dynamic display method of image multiple exposure fusion of the present invention.
  • FIG. 4 is a flowchart of another step of the high dynamic display method of image multiple exposure fusion of the present invention.
  • FIG. 5 is a schematic diagram of an original image of a high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 6 is a schematic diagram of a low-exposure image of an embodiment of an original image of a high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 7 is a schematic diagram of a medium-exposure image of an embodiment of the original image of the high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 8 is a schematic diagram of a high-exposure image of an embodiment of the original image of the high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 9 is a schematic diagram of an exposure image of another embodiment of the original image of the high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 10 is a schematic diagram of an image base layer of another embodiment of the original image of the high-dynamic display method of image multiple exposure fusion according to the present invention.
  • FIG. 11 is a schematic diagram of an image detail layer of another embodiment of the original image of the high-dynamic display method of image multiple exposure fusion according to the present invention.
  • the high dynamic display method of image multiple exposure fusion of the present invention can be executed by a computer 10, and the method includes: computer providing step S01, original image input step S02, multiple exposure image generating step S03 , Human eye region of interest information extraction step S04, human eye region of interest weight calculation step S05, base layer and detail layer extraction step S06, fusion image step S07, and output step S08.
  • the computer provides step S01, including providing the computer 10.
  • the computer 10 may be a computer, smart phone, tablet computer, smart watch, or the like.
  • the computer 10 at least includes a central processing unit 11, a memory 12, a storage 13, an input interface 14, and an output interface 15 that are electrically connected to each other.
  • the memory 12 may be a dynamic random access memory (DRAM).
  • the storage 13 may be a hard disk (Hard Disk) Drive, HDD) or Solid State Drive (SSD).
  • the input interface 14 may be an electrical connector, such as a universal serial bus (Universal Serial Bus, USB) electrical connector.
  • the output interface 15 may be an electrical connector, such as a high definition multimedia interface (High Definition Multimedia Interface, HDMI) connector for outputting images to external electronic devices, such as liquid crystal display panels.
  • HDMI High Definition Multimedia Interface
  • the original image input step S02 includes inputting the original image OG to the computer 10.
  • the original image OG may be any picture.
  • the original image OG is a landscape image and has characteristics such as sky, distant view, and close view.
  • the multi-exposure image generation step S03 described in the figure, including computing by the computer 10, and generating multiple exposure images L1, L2, L3 using an appropriate S-shaped function, as shown in FIG. .
  • the plurality of exposure images L1, L2, and L3 are grayscale images.
  • the exposure values of the plurality of exposure images L1, L2, L3 are different, and the plurality of exposure images L1, L2, L3 may be a low exposure image L1, a medium exposure image L2, and a high exposure image L3, respectively.
  • the human eye region of interest information extraction step S04 includes computing through the computer 10, and extracting multiple of each of the exposure images L1, L2, L3 through an image saliency model GBVS (Graph-based Visual Saliency) Regions of interest R1, R2, R3 of the personal eye.
  • GBVS Graph-based Visual Saliency
  • the lowest exposure human eye region of interest R1 is the sky bright region
  • the highest exposure human eye region of interest R3 is the darkest region of the near scene
  • the middle exposure human eye region of interest R2 is the distant region.
  • Wk (i, j) exp (- ⁇ x ((L (i, j) / (Lmed, k))-1) 2 )
  • the weight values W1, W2, W3 for the plurality of exposure images L1, L2, L3 are obtained.
  • the step S06 of extracting the base layer and the detail layer includes computing by the computer 10 while extracting the image base layers B1, B2 from each of the exposure images L1, L2, L3 using a Principal Component Analysis (PCA) method , B3 and image detail layers D1, D2, D3, as shown in FIG. 2; wherein the image detail layers D1, D2, D3 are obtained by subtracting the original image OG from the image base layers B1, B2, B3 obtain.
  • PCA Principal Component Analysis
  • the base layer and detail layer extraction step S06 includes a brightness extraction step, an image base layer generation step, and an image detail layer generation step.
  • the brightness extraction step includes performing calculations by the computer 10 to extract brightness values from the low-exposure image L1, the medium-exposure image L2, and the high-exposure image L3 generated from the original image OG, respectively, through PCA
  • the cumulative contribution rate of the first k eigenvalues is> 95%.
  • the step of generating the image base layer includes performing calculations by the computer 10 to reconstruct the image base layers B1, B2, and B3, which are original images, by PCA inverse transformation.
  • the step of generating the image detail layer includes computing the computer 10 to subtract the image base layers B1, B2, and B3 from the exposure images L1, L2, and L3 to obtain the image detail layers D1, D2, and D3.
  • FIGS. 9 to 11 are schematic diagrams of related images according to another embodiment of the present invention, wherein the image shown in FIG. 9 is an exposure image L of another original image OG, and the image shown in FIG. 10 is the exposure image L
  • the base layer B of FIG. 11 is the detail layer D of the exposure image L.
  • the step S07 of fusing images includes fusing all image base layers B1, B2, B3 to generate a fusing base layer CB, fusing all image detail layers D1, D2, D3 to generate a fusing detail layer CD, and finally fusing
  • the fusion base layer CB and the fusion detail layer CD are used to generate a fusion image CG.
  • the fusion image step S07 includes an image base layer fusion step S07a, an image detail layer fusion step S07b, and a fusion image generation step S07c.
  • the image basic layer fusion step S07a includes calculation by the computer 10 according to the weight value Wk (i, j) of each of the human eye interest regions in each of the exposure images L1, L2, L3. Multiple image base layers B1, B2, B3 are weighted and fused to generate a fused base layer CB.
  • the image basic layer fusion step S07a is based on a basic layer fusion calculation formula, and the basic fusion calculation formula is as follows:
  • the image detail layer fusion step S07b includes the operation of the computer 10 to fuse multiple image detail layers D1, D2, D3 to generate a fusion detail layer CD.
  • the image detail layer fusion step S07b includes first generating a plurality of detail enhancement coefficients a, b, and c, and then performing the plurality of image detail layers D1 through a detail layer fusion calculation formula D2, D3 fusion.
  • the detail enhancement coefficient a, b, or c is determined by the difference between the exposed images L1, L2, L3 of the original image OG and the image base layers B1, B2, B3. The greater the difference, the more details are enhanced.
  • the applied formula is as follows:
  • the detail enhancement coefficients a, b, or c 1 + sqrt (std (L) / 255-std (B) / 255); where L represents the gray value of the exposure images L1, L2, L3 of the original image OG, B represents the gray value of the base layer, sqrt represents the square root (Square Root), and Std represents the standard deviation (Standard Deviation).
  • L represents the gray value of the exposure images L1, L2, L3 of the original image OG
  • B represents the gray value of the base layer
  • sqrt represents the square root (Square Root)
  • Std represents the standard deviation (Standard Deviation).
  • the detail enhancement coefficients a, b, or c may also be referred to as detail adaptation coefficients.
  • D (a x D1 + b x D2 + c x D3) / (a + b + c); where D1, D2, and D3 respectively represent the gray values of the image detail layers D1, D2, and D3, and a, b, and c are detail enhancement coefficients.
  • the fusion image generating step S07c includes, through the operation of the computer 10, fusion of the fusion base layer CB and the fusion detail layer CD to generate a fusion image CG.
  • the fusion base layer CB and the fusion detail layer CD are fused through linear superposition.
  • the output step S08 includes outputting the fusion image CG to an external electronic device through the computer 10, for example, to a display.
  • the high dynamic display method of image multiple exposure fusion of the present invention generates multiple exposure images L1, L2, L3 with different exposure values from the original image OG, and then L2, L3 extract image base layers B1, B2, B3 and image detail layers D1, D2, D3, respectively, and then fuse the multiple image base layers B1, B2, B3 with specific weight values, and enhance detailed fusion with specific details
  • the multiple image detail layers D1, D2, D3, and finally the fusion base layer CB and the fusion detail layer CD are fused a second time to obtain a fused image CG.
  • the fusion image obtained by the above method of the present invention has been enhanced in detail for the human eye regions of interest R1, R2, R3. Therefore, compared with the prior art image enhancement method optimized according to the average value of the image, the present invention Has better display effect.
  • the method of the present invention has the following advantages:
  • the present invention performs region-based enhancement based on the HVS human eye perception system, respectively enhancing regions of human eye interest in different exposure images, and retaining more image details.
  • the PCA-based method of the present invention generates a base layer and a detail layer, which is faster than calculation using guide / bilateral filtering, and the effect is basically the same.
  • the method of adaptive detail enhancement coefficient of the present invention is better than the traditional method using an empirical magnification factor, which can balance the gap between the original image OG and the image detail layers D1, D2, D3, effectively avoiding noise amplification and insufficient detail enhancement to the image detail layer The impact of D1, D2, D3.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

本发明公开一种基于图像多重曝光融合的高动态显示方法，包括原图像输入步骤、多曝光图像生成步骤、人眼感兴趣区域资讯提取步骤、人眼感兴趣区域权重计算步骤、基础层及细节层提取步骤、以及融合图像步骤。本发明方法透过从原图像生成多个具有不同曝光值的曝光图像后，再对各所述曝光图像分别提取图像基础层及图像细节层，接着以特定权重值融合所述多个图像基础层，并以特定细节增强细数融合所述多个图像细节层，最后将融合基础层以及融合细节层进行第二次融合而得到融合图像，具有较佳的显示效果。

Description

基于图像多重曝光融合的高动态显示方法 技术领域
本发明是关于一种基于图像多重曝光融合的高动态显示方法，其对图像进行分区域增强，分别增强不同曝光图像人眼感兴趣区域，并在所述人眼感兴趣区域中保留更多细节，使图像整体视觉效果增强。
 背景技术
数位相机具有拍照快速、无须冲洗底片的优底，因此已取代传统相机多年。然而数位相机所拍摄的图像，高光容易过曝，低光容易过暗，因此有许多人发展出多曝光融合方法，将相机的图像进行数位后制处理，以得到高动态显示的图像。
现有技术多曝光融合方法通过在一张图像上建立一个合适的曝光函数来生成多张曝光图像。每张曝光图像的权重计算求取得是图像的均值作为中心值。
每个不同的曝光图像往往关注的重点不同，例如较暗的曝光图像往往关注的是最亮的那个区域（例如天空）。反之，最亮的曝光图像需要增强的地方是较暗的区域细节。因此，现有技术数统一的取均值并不能得到较好的图像显示效果，所后制出的图像往往发白或模糊。
故，有必要提供一种基于图像多重曝光融合的高动态显示方法，以解决现有技术所存在的问题。
 技术问题
本发明提供一种基于图像多重曝光融合的高动态显示方法，以解决现有技术的的图像经处理后发白或是模糊的问题。
 技术解决方案
本发明的主要目的在于提供一种多重曝光融合的高动态显示方法，包括：
多曝光图像生成步骤，包括使用合适的S型函数从原图像生成多张曝光图像；
人眼感兴趣区域资讯提取步骤，包括通过一基于图像显着性模型提取各所述曝光图像中的多个人眼感兴趣区域；
人眼感兴趣区域权重计算步骤，包括分别计算各所述曝光图像中的各所述人眼感兴趣区域的权重值；
基础层及细节层提取步骤，包括自各所述曝光图像中提取图像基础层及图像细节层；以及
融合图像步骤，包括融合所有图像基础层以生成融合基础层，融合所有图像细节层以生成融合细节层，最后融合所述融合基础层及所述融合细节层以生成融合图像。
在本发明的一实施例中，所述方法更包括：计算机提供步骤，包括提供计算机；原图像输入步骤，包括输入所述原图像到所述计算机；所述多曝光图像生成步骤、所述人眼感兴趣区域资讯提取步骤、所述人眼感兴趣区域权重计算步骤、所述基础层及细节层提取步骤、所述融合图像步骤是通过所述计算机运算而执行。
在本发明的一实施例中，所述多重曝光融合的高动态显示方法进一步包括输出步骤，所述输出步骤包括通过所述计算机，输出所述融合图像到外部电子装置。
在本发明的一实施例中，所述多个曝光图像的曝光值相异。
在本发明的一实施例中，所述多个曝光图像为灰度图。
在本发明的一实施例中，所述基础层及细节层提取步骤是使用主成分分析方法来自各所述曝光图像中提取所述图像基础层及所述图像细节层。
在本发明的一实施例中，所述融合图像步骤包括：图像基础层融合步骤，包括通过所述计算机运算，根据各所述曝光图像中各所述人眼感兴趣区域的权重值，对所述多个图像基础层进行加权融合，以生成所述融合基础层。
在本发明的一实施例中，所述融合图像步骤更包括：图像细节层融合步骤，包括通过所述计算机运算，对多个图像细节层进行融合，以生成所述融合细节层，其中，所述图像细节层融合步骤包括先产生多个细节增强系数，再透过一细节层融合计算式进行所述多个图像细节层的融合。
在本发明的一实施例中，所述融合图像步骤更包括：融合图像生成步骤，包括通过所述计算机运算，融合所述融合基础层以及所述融合细节层，以生成所述融合图像。
在本发明的一实施例中，所述细节增强系数= 1+sqrt(std(L)/255-std(B)/255)；其中L代表原图像之曝光图像的灰度值，B代表基础层之灰度值，sqrt代表平方根，Std代表标准差。
在本发明的一实施例中，所述细节层融合计算式如下：D = (a x D1+b x D2 + c x D3)/(a + b + c)；其中D1、D2、D3分别代表所述多个图像细节层的灰度值，a、b、c为所述细节增强系数。
在本发明的一实施例中，所述融合基础层以及所述融合细节层是透过线性迭加方式而融合成所述融合图像。
本发明的另一目的在于提供一种多重曝光融合的高动态显示方法，包括：
多曝光图像生成步骤，包括使用合适的S型函数从原图像生成多张曝光图像；
人眼感兴趣区域资讯提取步骤，包括通过一基于图像显着性模型提取各所述曝光图像中的多个人眼感兴趣区域；
人眼感兴趣区域权重计算步骤，包括分别计算各所述曝光图像中的各所述人眼感兴趣区域的权重值；
基础层及细节层提取步骤，包括自各所述曝光图像中提取图像基础层及图像细节层；以及
融合图像步骤，包括融合所有图像基础层以生成融合基础层，融合所有图像细节层以生成融合细节层，最后融合所述融合基础层及所述融合细节层以生成融合图像；
其中所述多个曝光图像的曝光值相异；
其中所述多个曝光图像为灰度图；
其中所述基础层及细节层提取步骤是使用主成分分析方法来自各所述曝光图像中提取所述图像基础层及所述图像细节层。
在本发明的一实施例中，所述方法更包括：计算机提供步骤，包括提供计算机；原图像输入步骤，包括输入所述原图像到所述计算机；以及所述多曝光图像生成步骤、所述人眼感兴趣区域资讯提取步骤、所述人眼感兴趣区域权重计算步骤、所述基础层及细节层提取步骤、所述融合图像步骤是通过所述计算机运算而执行。
在本发明的一实施例中，所述多重曝光融合的高动态显示方法进一步包括输出步骤，所述输出步骤包括通过所述计算机，输出所述融合图像到外部电子装置。
在本发明的一实施例中，所述融合图像步骤包括：图像基础层融合步骤，包括通过所述计算机运算，根据各所述曝光图像中各所述人眼感兴趣区域的权重值，对所述多个图像基础层进行加权融合，以生成所述融合基础层。
在本发明的一实施例中，所述融合图像步骤包括：图像细节层融合步骤，包括通过所述计算机运算，对多个图像细节层进行融合，以生成所述融合细节层，其中，所述图像细节层融合步骤包括先产生多个细节增强系数，再透过一细节层融合计算式进行所述多个图像细节层的融合。
在本发明的一实施例中，所述融合图像步骤包括：融合图像生成步骤，包括通过所述计算机运算，融合所述融合基础层以及所述融合细节层，以生成所述融合图像。
在本发明的一实施例中，所述细节增强系数= 1+sqrt(std(L)/255-std(B)/255)；其中L代表原图像之曝光图像的灰度值，B代表基础层之灰度值，sqrt代表平方根，Std代表标准差。
在本发明的一实施例中，所述细节层融合计算式如下：D = (a x D1+b x D2 + c x D3)/(a + b + c)；其中D1、D2、D3分别代表所述多个图像细节层的灰度值，a、b、c为所述细节增强系数。
 有益效果
相较于现有技术，本发明图像多重曝光融合的高动态显示方法透过从原图像生成多个具有不同曝光值的曝光图像后，再对各所述曝光图像分别提取图像基础层及图像细节层，接着以特定权重值融合所述多个图像基础层，并以特定细节增强细数融合所述多个图像细节层，最后将融合基础层以及融合细节层进行第二次融合而得到融合图像。透过上述本发明方法所得到的融合图像，已经针对人眼感兴趣区域进行细节的强化，因此相较于根据图像的平均值进行优化的现有技术图像强化方法，本发明具有较佳的显示效果。
为让本发明的上述内容能更明显易懂，下文特举优选实施例，并配合所附图式，作详细说明如下：
 附图说明
图1是本发明图像多重曝光融合的高动态显示方法所适用的计算机的架构方块图。
图2是本发明图像多重曝光融合的高动态显示方法的图像处理流程方块图。
图3是本发明图像多重曝光融合的高动态显示方法的步骤流程图。
图4是本发明图像多重曝光融合的高动态显示方法的另一步骤流程图。
图5是本发明图像多重曝光融合的高动态显示方法的原图像示意图。
图6是本发明图像多重曝光融合的高动态显示方法的原图像一实施例的低曝光图像示意图。
图7是本发明图像多重曝光融合的高动态显示方法的的原图像一实施例的中曝光图像示意图。
图8是本发明图像多重曝光融合的高动态显示方法的原图像一实施例的高曝光图像示意图。
图9是本发明图像多重曝光融合的高动态显示方法的原图像另一实施例的曝光图像示意图。
图10是本发明图像多重曝光融合的高动态显示方法的原图像另一实施例的图像基础层示意图。
图11是本发明图像多重曝光融合的高动态显示方法的原图像另一实施例的图像细节层示意图。
 本发明的实施方式
请参照图1以及图3，本发明图像多重曝光融合的高动态显示方法可透过一计算机执行10，且所述方法包括：计算机提供步骤S01、原图像输入步骤S02、多曝光图像生成步骤S03、人眼感兴趣区域资讯提取步骤S04、人眼感兴趣区域权重计算步骤S05、基础层及细节层提取步骤S06、融合图像步骤S07、以及输出步骤S08。
请参照图2及图3，所述计算机提供步骤S01，包括提供计算机10。所述计算机10可为电脑、智慧手机、平板电脑、智慧手表等等。此外，所述计算机10至少包括相互电连接的中央处理器11、记忆体12、储存器13、输入介面14以及输出介面15。所述记忆体12可为随机存取动态记忆体12(Dynamic Random-access Memory, DRAM)。所述储存器13可为硬碟(Hard Disk Drive, HDD)或是固态硬碟(Solid State Drive, SSD)。所述输入介面14可为电连接器，例如通用序列汇流排(Universal Serial Bus, USB)电连接器。所述输出介面15可为电连接器，例如高画质多媒体介面(High Definition Multimedia Interface, HDMI)连接器，用于输出图像到外部电子装置，例如液晶显示面板。
请参照图5，所述原图像输入步骤S02，包括输入原图像OG到所述计算机10。在本发明实施例中，所述原图像OG可为任何图片，在如图3所示的本发明一实施例中，所述原图像OG为一风景图而具有天空、远景、近景等特征。
请参照图6至图8，请参照图所述多曝光图像生成步骤S03，包括通过所述计算机10运算，使用合适的S型函数生成多张曝光图像L1、L2、L3，如图2所示。所述多个曝光图像L1、L2、L3为灰度(Grayscale)图。所述多个曝光图像L1、L2、L3的曝光值相异，且所述多个曝光图像L1、L2、L3可分别为低曝光图像L1、中曝光图像L2及高曝光图像L3。
所述人眼感兴趣区域资讯提取步骤S04，包括通过所述计算机10运算，通过一基于图像显着性模型GBVS(Graph-based Visual Saliency)提取各所述曝光图像L1、L2、L3中的多个人眼感兴趣区域R1、R2、R3。
其中，最低曝光人眼感兴趣区域R1为天空明亮区域；最高曝光人眼感兴趣区域R3为近景最暗区域；中曝光人眼感兴趣区域R2为远景区域。
所述人眼感兴趣区域权重计算步骤S05，包括分别计算各所述曝光图像L1、L2、L3的区域均值Lmed 1, 2, 3, 4，并建立高斯权重函数 k=1, 2, 3，最后分别计算各所述曝光图像L1、L2、L3中的各所述人眼感兴趣区域的权重值Wk(i,j)如下：
Wk(i,j) = exp(-α x ((L(i,j)/(Lmed, k))-1) 2)
根据上述计算式，得出对所述多个曝光图像L1、L2、L3的权重值W1、W2、W3。
所述基础层及细节层提取步骤S06，包括通过所述计算机10运算，同时使用主成分分析 (Principal Component Analysis, PCA)方法自各所述曝光图像L1、L2、L3中提取图像基础层B1、B2、B3及图像细节层D1、D2、D3，如图2所示；其中所述图像细节层D1、D2、D3是由所述原图像OG与所述图像基础层B1、B2、B3相减而获得。所述PCA方法为本领域技术人员所能了解的技术，于本文中不再赘述。
所述基础层及细节层提取步骤S06包括亮度提取步骤、图像基础层生成步骤、以及图像细节层生成步骤。
所述亮度提取步骤，包括通过所述计算机10运算，从所述原图像OG生成的所述低曝光图像L1、所述中曝光图像L2、所述高曝光图像L3中分别提取亮度值，经过PCA降维，取前k个特征值累积贡献率>95%。
所述图像基础层生成步骤，包括通过所述计算机10运算，以PCA逆变换重构出原图即图像基础层B1、B2、B3。
所述图像细节层生成步骤，包括通过所述计算机10运算，将各所述曝光图像L1、L2、L3减去图像基础层B1、B2、B3，得到图像细节层D1、D2、D3。
请参照图9至图11，为本发明另一实施例的相关图像的示意图，其中，图9所示者为另一原图像OG的曝光图像L，图10所示者为所述曝光图像L的基础层B，图11所示者为所述曝光图像L的细节层D。
请参照图4，所述融合图像步骤S07，包括融合所有图像基础层B1、B2、B3以生成融合基础层CB，融合所有图像细节层D1、D2、D3以生成融合细节层CD，最后融合所述融合基础层CB及所述融合细节层CD以生成融合图像CG。于发明较佳实施例中，所述融合图像步骤S07包括图像基础层融合步骤S07a、图像细节层融合步骤S07b、以及融合图像生成步骤S07c。
所述图像基础层融合步骤S07a，包括通过所述计算机10运算，根据各所述曝光图像L1、L2、L3中各所述人眼感兴趣区域的权重值Wk(i,j)，对所述多个图像基础层B1、B2、B3进行加权融合，以生成一融合基础层CB。于本发明较佳实施例中，所述图像基础层融合步骤S07a是以一基础层融合计算式进行融合，所述基础融合计算式如下：
L = B1 x W1+ B2 x W2+ B3 x W3。
所述图像细节层融合步骤S07b，包括通过所述计算机10运算，对多个图像细节层D1、D2、D3进行融合，以生成一融合细节层CD。于本发明较佳实施例中，所述图像细节层融合步骤S07b包括先产生多个细节增强系数a、b、c，再透过一细节层融合计算式进行所述多个图像细节层D1、D2、D3的融合。
细节增强系数a、b、或c，由原图像OG的曝光图像L1、L2、L3与图像基础层B1、B2、B3差异决定，差异越大，细节增强越多。所应用的公式如下：
所述细节增强系数a、b、或c = 1+sqrt(std(L)/255-std(B)/255)；其中L代表原图像OG之曝光图像L1、L2、L3的灰度值，B代表基础层之灰度值，sqrt代表平方根(Square Root)，Std代表标准差(Standard Deviation)。此外，所述细节增强系数a、b、或c也可称为细节自适应系数。
所述细节层融合计算式如下：
D = (a x D1+b x D2 + c x D3)/(a + b + c)；其中D1、D2、D3分别代表图像细节层D1、D2、D3的灰度值，a、b、c为细节增强系数。
所述融合图像生成步骤S07c，包括通过所述计算机10运算，融合所述融合基础层CB以及所述融合细节层CD，以生成一融合图像CG。于本发明较佳实施例中，所述融合基础层CB以及所述融合细节层CD是透过线性迭加方式而融合。
所述输出步骤S08，包括通过所述计算机10，输出所述融合图像CG到外部电子装置，例如输出到显示器。
相较于现有技术，本发明图像多重曝光融合的高动态显示方法透过从原图像OG生成多个具有不同曝光值的曝光图像L1、L2、L3后，再对各所述曝光图像L1、L2、L3分别提取图像基础层B1、B2、B3及图像细节层D1、D2、D3，接着以特定权重值融合所述多个图像基础层B1、B2、B3，并以特定细节增强细数融合所述多个图像细节层D1、D2、D3，最后将融合基础层CB以及融合细节层CD进行第二次融合而得到融合图像CG。透过上述本发明方法所得到的融合图像，已经针对人眼感兴趣区域R1、R2、R3进行细节的强化，因此相较于根据图像的平均值进行优化的现有技术图像强化方法，本发明具有较佳的显示效果。此外，本发明方法尚且具下列优点：
1.本发明基于HVS人眼感知系统进行分区域增强，分别增强不同曝光图像人眼感兴趣区域，保留更多图像细节。
2.本发明基于PCA的方法生成基础层和细节层，比使用引导/双边滤波计算更加快速，效果基本相同。
3.本发明自适应细节增强系数的方法比传统方法使用一个经验放大系数更优，能平衡原图像OG与图像细节层D1、D2、D3差距，有效避免噪声放大及细节增强不足对图像细节层D1、D2、D3的影响。

Claims (20)

  1. 一种多重曝光融合的高动态显示方法，包括：
    多曝光图像生成步骤，包括使用合适的S型函数从原图像生成多张曝光图像；
    人眼感兴趣区域资讯提取步骤，包括通过一基于图像显着性模型提取各所述曝光图像中的多个人眼感兴趣区域；
    人眼感兴趣区域权重计算步骤，包括分别计算各所述曝光图像中的各所述人眼感兴趣区域的权重值；
    基础层及细节层提取步骤，包括自各所述曝光图像中提取图像基础层及图像细节层；以及
    融合图像步骤，包括融合所有图像基础层以生成融合基础层，融合所有图像细节层以生成融合细节层，最后融合所述融合基础层及所述融合细节层以生成融合图像。
  2. 如权利要求1所述的多重曝光融合的高动态显示方法，其中：
    所述方法更包括：
    计算机提供步骤，包括提供计算机；
    原图像输入步骤，包括输入所述原图像到所述计算机；以及
    所述多曝光图像生成步骤、所述人眼感兴趣区域资讯提取步骤、所述人眼感兴趣区域权重计算步骤、所述基础层及细节层提取步骤、所述融合图像步骤是通过所述计算机运算而执行。
  3. 如权利要求2所述的多重曝光融合的高动态显示方法，其中所述多重曝光融合的高动态显示方法进一步包括输出步骤，所述输出步骤包括通过所述计算机，输出所述融合图像到外部电子装置。
  4. 如权利要求1所述的多重曝光融合的高动态显示方法，其中所述多个曝光图像的曝光值相异。
  5. 如权利要求1所述的多重曝光融合的高动态显示方法，其中所述多个曝光图像为灰度图。
  6. 如权利要求1所述的多重曝光融合的高动态显示方法，其中所述基础层及细节层提取步骤是使用主成分分析方法来自各所述曝光图像中提取所述图像基础层及所述图像细节层。
  7. 如权利要求2所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤包括：图像基础层融合步骤，包括通过所述计算机运算，根据各所述曝光图像中各所述人眼感兴趣区域的权重值，对所述多个图像基础层进行加权融合，以生成所述融合基础层。
  8. 如权利要求7所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤更包括：图像细节层融合步骤，包括通过所述计算机运算，对多个图像细节层进行融合，以生成所述融合细节层，其中，所述图像细节层融合步骤包括先产生多个细节增强系数，再透过一细节层融合计算式进行所述多个图像细节层的融合。
  9. 如权利要求8所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤更包括：融合图像生成步骤，包括通过所述计算机运算，融合所述融合基础层以及所述融合细节层，以生成所述融合图像。
  10. 如权利要求8所述的多重曝光融合的高动态显示方法，其中所述细节增强系数= 1+sqrt(std(L)/255-std(B)/255)；其中L代表原图像之曝光图像的灰度值，B代表基础层之灰度值，sqrt代表平方根，Std代表标准差。
  11. 如权利要求8所述的多重曝光融合的高动态显示方法，其中所述细节层融合计算式如下：D = (a x D1+b x D2 + c x D3)/(a + b + c)；其中D1、D2、D3分别代表所述多个图像细节层的灰度值，a、b、c为所述细节增强系数。
  12. 如权利要求1所述的多重曝光融合的高动态显示方法，其中所述融合基础层以及所述融合细节层是透过线性迭加方式而融合成所述融合图像。
  13. 一种多重曝光融合的高动态显示方法，包括：
    多曝光图像生成步骤，包括使用合适的S型函数从原图像生成多张曝光图像；
    人眼感兴趣区域资讯提取步骤，包括通过一基于图像显着性模型提取各所述曝光图像中的多个人眼感兴趣区域；
    人眼感兴趣区域权重计算步骤，包括分别计算各所述曝光图像中的各所述人眼感兴趣区域的权重值；
    基础层及细节层提取步骤，包括自各所述曝光图像中提取图像基础层及图像细节层；以及
    融合图像步骤，包括融合所有图像基础层以生成融合基础层，融合所有图像细节层以生成融合细节层，最后融合所述融合基础层及所述融合细节层以生成融合图像；
    其中所述多个曝光图像的曝光值相异；
    其中所述多个曝光图像为灰度图；
    其中所述基础层及细节层提取步骤是使用主成分分析方法来自各所述曝光图像中提取所述图像基础层及所述图像细节层。
  14. 如权利要求13所述的多重曝光融合的高动态显示方法，其中：
    所述方法更包括：
    计算机提供步骤，包括提供计算机；
    原图像输入步骤，包括输入所述原图像到所述计算机；以及
    所述多曝光图像生成步骤、所述人眼感兴趣区域资讯提取步骤、所述人眼感兴趣区域权重计算步骤、所述基础层及细节层提取步骤、所述融合图像步骤是通过所述计算机运算而执行。
  15. 如权利要求14所述的多重曝光融合的高动态显示方法，其中所述多重曝光融合的高动态显示方法进一步包括输出步骤，所述输出步骤包括通过所述计算机，输出所述融合图像到外部电子装置。
  16. 如权利要求14所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤包括：图像基础层融合步骤，包括通过所述计算机运算，根据各所述曝光图像中各所述人眼感兴趣区域的权重值，对所述多个图像基础层进行加权融合，以生成所述融合基础层。
  17. 如权利要求16所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤更包括：图像细节层融合步骤，包括通过所述计算机运算，对多个图像细节层进行融合，以生成所述融合细节层，其中，所述图像细节层融合步骤包括先产生多个细节增强系数，再透过一细节层融合计算式进行所述多个图像细节层的融合。
  18. 如权利要求17所述的多重曝光融合的高动态显示方法，其中所述融合图像步骤包括：融合图像生成步骤，包括通过所述计算机运算，融合所述融合基础层以及所述融合细节层，以生成所述融合图像。
  19. 如权利要求17所述的多重曝光融合的高动态显示方法，其中所述细节增强系数 = 1+sqrt(std(L)/255-std(B)/255)；其中L代表原图像之曝光图像的灰度值，B代表基础层之灰度值，sqrt代表平方根，Std代表标准差。
  20. 如权利要求17所述的多重曝光融合的高动态显示方法，其中所述细节层融合计算式如下：D = (a x D1+b x D2 + c x D3)/(a + b + c)；其中D1、D2、D3分别代表所述多个图像细节层的灰度值，a、b、c为所述细节增强系数。
PCT/CN2019/072435 2018-11-09 2019-01-18 基于图像多重曝光融合的高动态显示方法 Ceased WO2020093600A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811332568.8A CN109636765B (zh) 2018-11-09 2018-11-09 基于图像多重曝光融合的高动态显示方法
CN201811332568.8 2018-11-09

Publications (1)

Publication Number Publication Date
WO2020093600A1 true WO2020093600A1 (zh) 2020-05-14

Family

ID=66067647

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/072435 Ceased WO2020093600A1 (zh) 2018-11-09 2019-01-18 基于图像多重曝光融合的高动态显示方法

Country Status (2)

Country Link
CN (1) CN109636765B (zh)
WO (1) WO2020093600A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112288664A (zh) * 2020-09-25 2021-01-29 北京迈格威科技有限公司 高动态范围图像的融合方法、装置和电子设备
CN113610861A (zh) * 2021-06-21 2021-11-05 重庆海尔制冷电器有限公司 制冷设备内食材图像处理方法、制冷设备及可读存储介质
CN113628141A (zh) * 2021-08-18 2021-11-09 上海磐启微电子有限公司 一种基于高低曝光图像融合的hdr细节增强方法
CN117061841A (zh) * 2023-06-12 2023-11-14 深圳市博盛医疗科技有限公司 一种双晶片内窥镜成像方法及成像装置
CN119624846A (zh) * 2024-12-10 2025-03-14 深圳市互通创新科技有限公司 一种用于监控设备的图像特征增强方法及装置

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110087003B (zh) * 2019-04-30 2021-03-23 Tcl华星光电技术有限公司 多曝光图像融合方法
CN110602384B (zh) * 2019-08-27 2022-03-29 维沃移动通信有限公司 曝光控制方法及电子设备
CN111898532A (zh) * 2020-07-30 2020-11-06 杭州海康威视数字技术股份有限公司 一种图像处理方法、装置、电子设备及监控系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105279746A (zh) * 2014-05-30 2016-01-27 西安电子科技大学 一种基于双边滤波的多曝光图像融合方法
CN106815827A (zh) * 2017-01-18 2017-06-09 聚龙智瞳科技有限公司 基于Bayer格式的图像融合方法及图像融合装置
CN107220956A (zh) * 2017-04-18 2017-09-29 天津大学 一种基于多幅具有不同曝光度的ldr图像的hdr图像融合方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101633893B1 (ko) * 2010-01-15 2016-06-28 삼성전자주식회사 다중노출 영상을 합성하는 영상합성장치 및 방법
KR20130031574A (ko) * 2011-09-21 2013-03-29 삼성전자주식회사 영상 처리 방법 및 영상 처리 장치
CN103247036B (zh) * 2012-02-10 2016-05-18 株式会社理光 多曝光图像融合方法和装置
CN104077759A (zh) * 2014-02-28 2014-10-01 西安电子科技大学 一种基于色觉感知及全局质量因子的多曝光度图像融合方法
CN105809641B (zh) * 2016-03-09 2018-02-16 北京理工大学 一种去雾图像的曝光补偿和边缘增强方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105279746A (zh) * 2014-05-30 2016-01-27 西安电子科技大学 一种基于双边滤波的多曝光图像融合方法
CN106815827A (zh) * 2017-01-18 2017-06-09 聚龙智瞳科技有限公司 基于Bayer格式的图像融合方法及图像融合装置
CN107220956A (zh) * 2017-04-18 2017-09-29 天津大学 一种基于多幅具有不同曝光度的ldr图像的hdr图像融合方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
SON, C. H. ET AL.: "Layer-based Approach for Image Pair Fusion", IEEE TRANSACTIONS ON IMAGE PROCESSING, vol. 25, no. 6, 1 June 2016 (2016-06-01), pages 2866 - 2881, XP011611545, ISSN: 1057-7149, DOI: 10.1109/TIP.2016.2556618 *
ZHANG, S ET AL: "A Generating Method of High Dynamic Range Images Using Principal Component Analysis and Gradient Pyramid", JOURNAL OF XI'AN JIAOTONG UNIVERSITY, vol. 52, no. 4, 1 April 2018 (2018-04-01), pages 150 - 157, XP055706217, ISSN: 0253-987X, DOI: 10.7652/xjtub201804022 *

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112288664A (zh) * 2020-09-25 2021-01-29 北京迈格威科技有限公司 高动态范围图像的融合方法、装置和电子设备
CN113610861A (zh) * 2021-06-21 2021-11-05 重庆海尔制冷电器有限公司 制冷设备内食材图像处理方法、制冷设备及可读存储介质
CN113610861B (zh) * 2021-06-21 2023-11-14 重庆海尔制冷电器有限公司 制冷设备内食材图像处理方法、制冷设备及可读存储介质
CN113628141A (zh) * 2021-08-18 2021-11-09 上海磐启微电子有限公司 一种基于高低曝光图像融合的hdr细节增强方法
CN113628141B (zh) * 2021-08-18 2023-11-28 上海磐启微电子有限公司 一种基于高低曝光图像融合的hdr细节增强方法
CN117061841A (zh) * 2023-06-12 2023-11-14 深圳市博盛医疗科技有限公司 一种双晶片内窥镜成像方法及成像装置
CN119624846A (zh) * 2024-12-10 2025-03-14 深圳市互通创新科技有限公司 一种用于监控设备的图像特征增强方法及装置

Also Published As

Publication number Publication date
CN109636765B (zh) 2021-04-02
CN109636765A (zh) 2019-04-16

Similar Documents

Publication Publication Date Title
WO2020093600A1 (zh) 基于图像多重曝光融合的高动态显示方法
Lee et al. Deep chain hdri: Reconstructing a high dynamic range image from a single low dynamic range image
Wang et al. Pseudo-multiple-exposure-based tone fusion with local region adjustment
CN111445564B (zh) 人脸纹理图像生成方法、装置、计算机设备和存储介质
CN112602088B (zh) 提高弱光图像的质量的方法、系统和计算机可读介质
CN106920221B (zh) 兼顾亮度分布和细节呈现的曝光融合方法
US20140085398A1 (en) Real-time automatic scene relighting in video conference sessions
CN112541877A (zh) 基于条件生成对抗网络的去模糊方法、系统、设备及介质
CN111047543B (zh) 图像增强方法、装置和存储介质
US20240296531A1 (en) System and methods for depth-aware video processing and depth perception enhancement
CN111327887A (zh) 电子装置及其操作方法，以及处理电子装置的图像的方法
CN105989574A (zh) 图像处理装置及图像景深处理方法
CN107147851B (zh) 照片处理方法、装置、计算机可读存储介质及电子设备
CN110428362A (zh) 图像hdr转换方法及装置、存储介质
US20230162335A1 (en) Image adjustment apparatus, image adjustment method and program
CN112153240A (zh) 一种调整图像画质方法、装置及可读存储介质
CN115482159A (zh) 图像增强方法和设备
CN115797224A (zh) 一种去除鬼影的高动态图像生成方法、装置及存储介质
CN110473156B (zh) 图像信息的处理方法、装置、存储介质及电子设备
US10915996B2 (en) Enhancement of edges in images using depth information
US11360555B2 (en) Systems and methods for automatic eye gaze refinement
Lo et al. High dynamic range (hdr) video image processing for digital glass
CN112651911A (zh) 一种基于偏振图像的高动态范围成像生成方法
CN111836058B (zh) 用于实时视频播放方法、装置、设备以及存储介质
CN105450943B (zh) 产生图像散景效果的方法及图像获取装置

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: 19881473

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19881473

Country of ref document: EP

Kind code of ref document: A1