CN110415187B - Image processing method and image processing system - Google Patents
Image processing method and image processing system Download PDFInfo
- Publication number
- CN110415187B CN110415187B CN201910610059.5A CN201910610059A CN110415187B CN 110415187 B CN110415187 B CN 110415187B CN 201910610059 A CN201910610059 A CN 201910610059A CN 110415187 B CN110415187 B CN 110415187B
- Authority
- CN
- China
- Prior art keywords
- image
- transition
- neural network
- transition image
- size
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 17
- 230000007704 transition Effects 0.000 claims abstract description 141
- 238000013528 artificial neural network Methods 0.000 claims abstract description 70
- 239000011159 matrix material Substances 0.000 claims abstract description 11
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 230000035772 mutation Effects 0.000 abstract description 6
- 230000010339 dilation Effects 0.000 description 26
- 238000005516 engineering process Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000000034 method Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000002087 whitening effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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; CALCULATING OR 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; 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
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides an image processing method and an image processing system. The image processing method comprises the following steps: receiving an original image, and reducing the size of the original image to obtain a first transition image; providing a first expansion convolution neural network, and performing expansion convolution processing on the first transition image by using the first expansion convolution neural network so as to enhance the image quality of the first transition image and obtain a second transition image; amplifying the size of the second transition image to obtain a third transition image, wherein the size of the third transition image is the same as that of the original image; splicing the original image and the third transition image matrix to obtain a fourth transition image; and providing a second expansion convolution neural network, and performing expansion convolution processing on the fourth transition image by using the second expansion convolution neural network so as to recover the detail information of the fourth transition image and obtain the target image. The invention can reduce the operation amount of the neural network under the condition of not reducing the detail information of the image, and avoid local mutation of the processed image.
Description
Technical Field
The present invention relates to the field of display technologies, and in particular, to an image processing method and an image processing system.
Background
With the development of science and technology, people have higher and higher requirements on display devices, and the display devices become the development direction and targets of televisions in a lighter, clearer and more vivid way. Besides the modes of improving the television resolution on hardware, increasing the dynamic range, the color gamut range and the like, the image processing system arranged in the display device can be used for commanding and coordinating various functions of the display device to optimize image signals one by one, thereby bringing better image quality to users on the basis of the existing hardware,
the image enhancement technology is one of image processing technologies, and can remarkably improve the image quality, so that the image content is more layered and the subjective observation effect is more in line with the requirements of people. In real life, the original image often has various problems, such as: when in photographing, the aperture is small, so that the image is dark; the contrast of the scene is low, so that the image emphasis is not prominent; overexposure can cause image disorders and whitening of the picture. The problems can be effectively solved through an image enhancement technology, and the display quality is improved.
Common image enhancement techniques include: saturation enhancement and contrast enhancement, contrast enhancement is of higher interest than saturation enhancement. The contrast enhancement is to increase the distribution range of image gray scale by adjusting the gray scale distribution of the image, so as to improve the contrast of the whole or part of the image and improve the visual effect.
Neural network architecture is a mathematical model that applies a structure similar to brain neurosynaptic connections for information processing. It has the greatest advantage of being able to be used as a mechanism for approximating an arbitrary function, "learning" from observed data. In recent years, neural network architectures have been increasingly used in image processing algorithms, typically for example, expanding convolutional neural networks.
The expansion convolution neural network can maintain a small number of operation times while expanding the visual field of the neural network due to the characteristics of expansion convolution, and the neural network output of the original pixel precision is realized.
Disclosure of Invention
The invention aims to provide an image processing method which can reduce the calculation amount of a neural network, avoid local mutation of an enhanced image and not reduce the detail information of the enhanced image.
The present invention is also directed to an image processing system, which can reduce the computation amount of a neural network, avoid local abrupt change of an enhanced image, and not reduce the detail information of the enhanced image.
In order to achieve the above object, the present invention provides an image processing method, comprising the steps of:
step S1, receiving an original image, and reducing the size of the original image to obtain a first transition image;
step S2, providing a first expansion convolution neural network, and performing expansion convolution processing on the first transition image by using the first expansion convolution neural network so as to enhance the image quality of the first transition image and obtain a second transition image;
step S3, enlarging the size of the second transition image to obtain a third transition image, wherein the size of the third transition image is the same as that of the original image;
step S4, splicing the original image and the third transition image matrix to obtain a fourth transition image;
and step S5, providing a second expansion convolution neural network, and performing expansion convolution processing on the fourth transition image by using the second expansion convolution neural network to recover the detail information of the fourth transition image to obtain the target image.
The first transition image has a size less than or equal to 256 pixels by 256 pixels.
The first dilation convolution neural network performs at least five-layer dilation convolution processing on the first transition image.
And the second expansion convolution neural network performs at least two layers of expansion convolution processing on the fourth transition image.
The size of the original image is reduced by the bilinear interpolation algorithm in step S1, and the size of the second transition image is enlarged by the bilinear interpolation algorithm in step S3.
The invention also provides an image processing system, which comprises an input unit, a reducing unit connected with the input unit, a first expansion convolution neural network connected with the reducing unit, an amplifying unit connected with the first expansion convolution neural network, a splicing unit connected with the amplifying unit and the input unit, a second expansion convolution neural network connected with the splicing unit and an output unit connected with the second expansion convolution neural network;
the input unit is used for receiving the original image and providing the original image to the reducing unit and the splicing unit;
the reducing unit is used for reducing the size of the original image to obtain a first transition image;
the first expansion convolution neural network is used for performing expansion convolution processing on the first transition image so as to enhance the image quality of the first transition image and obtain a second transition image;
the amplifying unit is used for amplifying the size of the second transition image to obtain a third transition image, and the size of the third transition image is the same as that of the original image;
the splicing unit is used for performing matrix splicing on the original image and the third transition image to obtain a fourth transition image;
the second expansion convolution neural network is used for performing expansion convolution processing on the fourth transition image so as to recover the detail information of the fourth transition image and obtain a target image;
the output unit is used for outputting a target image.
The first transition image has a size less than or equal to 256 pixels by 256 pixels.
The first dilation convolution neural network performs at least five-layer dilation convolution processing on the first transition image.
And the second expansion convolution neural network performs at least two layers of expansion convolution processing on the fourth transition image.
The reduction unit reduces the size of the original image by a bilinear interpolation algorithm, and the enlargement unit enlarges the size of the second transition image by the bilinear interpolation algorithm.
The invention has the beneficial effects that: the invention provides an image processing method, which comprises the following steps: receiving an original image, and reducing the size of the original image to obtain a first transition image; providing a first expansion convolution neural network, and performing expansion convolution processing on the first transition image by using the first expansion convolution neural network so as to enhance the image quality of the first transition image and obtain a second transition image; enlarging the size of the second transition image to obtain a third transition image, wherein the size of the third transition image is the same as that of the original image; splicing the original image and the third transition image matrix to obtain a fourth transition image; and providing a second expansion convolution neural network, performing expansion convolution processing on the fourth transition image by using the second expansion convolution neural network to recover the detail information of the fourth transition image to obtain a target image, reducing the calculation amount of the neural network under the condition of not reducing the detail information of the image, and avoiding local mutation of the processed image. The invention also provides an image processing system which can reduce the operation amount of the neural network under the condition of not reducing the detail information of the image and avoid local mutation of the processed image.
Drawings
For a better understanding of the nature and technical aspects of the present invention, reference should be made to the following detailed description of the invention, taken in conjunction with the accompanying drawings, which are provided for purposes of illustration and description and are not intended to limit the invention.
In the drawings, there is shown in the drawings,
FIG. 1 is a flow chart of an image processing method of the present invention;
FIG. 2 is a schematic diagram of an image change of the image processing method of the present invention;
FIG. 3 is a schematic diagram of an image processing system according to the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention, the following detailed description is given with reference to the preferred embodiments of the present invention and the accompanying drawings.
Referring to fig. 1 and fig. 2, the present invention provides an image processing method, including the following steps:
step S1, receiving an original image 1, and reducing the size of the original image 1 to obtain a first transition image 2;
specifically, the size of the first transition image 2 is less than or equal to 256 pixels × 256 pixels, and preferably, the size of the first transition image 2 is equal to 256 pixels × 256 pixels.
Specifically, in step S1, the size of the original image 1 is reduced by a bilinear interpolation algorithm.
Step S2, providing a first dilation convolution neural network 30, and performing dilation convolution processing on the first transition image 2 by using the first dilation convolution neural network 30 to enhance the image quality of the first transition image, so as to obtain a second transition image 3.
Specifically, the first dilated convolution neural network 30 performs at least five-layer dilated convolution processing on the first transition image 2, and preferably, the first dilated convolution neural network 30 performs five-layer dilated convolution processing on the first transition image 2.
And step S3, enlarging the size of the second transition image 3 to obtain a third transition image 4, where the size of the third transition image 4 is the same as the size of the original image 1.
Specifically, the size of the second transition image 3 is enlarged by the bilinear interpolation algorithm in step S3.
And step S4, performing matrix splicing on the original image 1 and the third transition image 4 to obtain a fourth transition image 5.
Step S5, providing a second dilation convolution neural network 60, and performing dilation convolution processing on the fourth transition image 5 by using the second dilation convolution neural network 60 to recover the detail information of the fourth transition image 5, so as to obtain the target image 6.
Specifically, the second dilation convolution neural network 60 performs at least two layers of dilation convolution processing on the fourth transition image 5, and preferably, the second dilation convolution neural network 60 performs two layers of dilation convolution processing on the fourth transition image 5.
It should be noted that, in the present invention, the original image 1 is reduced to the first transition image 2 with a fixed size, and image enhancement is performed on the first transition image 2 with a smaller size, so as to ensure that the first dilation convolution neural network 30 has a sufficient field of view to cover the whole image all the time in the enhancement process, thereby achieving that the original images with different sizes have sufficient field of view when image enhancement is performed, avoiding local sudden change of the image caused by insufficient field of view, greatly reducing the amount of operation of the first dilation convolution neural network 30, and reducing the network cost.
Further, the invention amplifies the small-sized second transition image 3 enhanced by the first expansion convolution neural network 30 to obtain a third transition image 4 with the same size as the original image, performs matrix splicing on the third transition image 4 and the original image 1 through transition connection to restore the size of the image, and finally restores image detail information through the second expansion convolution neural network 30 to obtain an enhanced target image 6, wherein the size of the target image 6 is the same as that of the original image 1.
Referring to fig. 3 and fig. 2, the present invention further provides an image processing system, which includes an input unit 10, a reducing unit 20 connected to the input unit 10, a first convolutional neural network 30 connected to the reducing unit 20, an amplifying unit 40 connected to the first convolutional neural network 30, a splicing unit 50 connected to the amplifying unit 40 and the input unit 10, a second convolutional neural network 60 connected to the splicing unit 50, and an output unit 70 connected to the second convolutional neural network 60;
the input unit 10 is configured to receive an original image 1, and provide the original image 1 to the reduction unit 20 and the stitching unit 50;
the reducing unit 20 is configured to reduce the size of the original image 1 to obtain a first transition image 2;
the first dilation convolution neural network 30 is configured to perform dilation convolution processing on the first transition image 2 to enhance the image quality of the first transition image, so as to obtain a second transition image 3;
the enlarging unit 40 is configured to enlarge the size of the second transition image 3 to obtain a third transition image 4, where the size of the third transition image 4 is the same as the size of the original image 1;
the splicing unit 50 is configured to perform matrix splicing on the original image 1 and the third transition image 4 to obtain a fourth transition image 5;
the second dilation convolution neural network 60 is configured to perform dilation convolution processing on the fourth transition image 5 to recover detail information of the fourth transition image 5, so as to obtain a target image 6;
the output unit 70 is used for outputting the target image 6.
Specifically, the size of the first transition image 2 is less than or equal to 256 pixels × 256 pixels, and preferably, the size of the first transition image 2 is equal to 256 pixels × 256 pixels
Specifically, the first dilated convolution neural network 30 performs at least five-layer dilated convolution processing on the first transition image 2, and preferably, the first dilated convolution neural network 30 performs five-layer dilated convolution processing on the first transition image 2
Specifically, the second dilation convolution neural network 60 performs at least two layers of dilation convolution processing on the fourth transition image 5, and preferably, the second dilation convolution neural network 60 performs two layers of dilation convolution processing on the fourth transition image 5.
Specifically, the reduction unit 20 reduces the size of the original image 1 by a bilinear interpolation algorithm, and the enlargement unit 40 enlarges the size of the second transition image 3 by the bilinear interpolation algorithm.
It should be noted that, in the present invention, the original image 1 is reduced to the first transition image 2 with a fixed size, and image enhancement is performed on the first transition image 2 with a smaller size, so as to ensure that the first dilation convolution neural network 30 has a sufficient field of view to cover the whole image all the time in the enhancement process, thereby achieving that the original images with different sizes have sufficient field of view when image enhancement is performed, avoiding local sudden change of the image caused by insufficient field of view, greatly reducing the amount of operation of the first dilation convolution neural network 30, and reducing the network cost.
Further, the invention amplifies the small-sized second transition image 3 enhanced by the first expansion convolution neural network 30 to obtain a third transition image 4 with the same size as the original image, performs matrix splicing on the third transition image 4 and the original image 1 through transition connection to restore the size of the image, and finally restores image detail information through the second expansion convolution neural network 30 to obtain an enhanced target image 6, wherein the size of the target image 6 is the same as that of the original image 1.
In summary, the present invention provides an image processing method, including the following steps: receiving an original image, and reducing the size of the original image to obtain a first transition image; providing a first expansion convolution neural network, and performing expansion convolution processing on the first transition image by using the first expansion convolution neural network so as to enhance the image quality of the first transition image and obtain a second transition image; enlarging the size of the second transition image to obtain a third transition image, wherein the size of the third transition image is the same as that of the original image; splicing the original image and the third transition image matrix to obtain a fourth transition image; and providing a second expansion convolution neural network, performing expansion convolution processing on the fourth transition image by using the second expansion convolution neural network to recover the detail information of the fourth transition image to obtain a target image, reducing the calculation amount of the neural network under the condition of not reducing the detail information of the image, and avoiding local mutation of the processed image. The invention also provides an image processing system which can reduce the operation amount of the neural network under the condition of not reducing the detail information of the image and avoid local mutation of the processed image.
As described above, it will be apparent to those skilled in the art that other various changes and modifications may be made based on the technical solution and concept of the present invention, and all such changes and modifications are intended to fall within the scope of the appended claims.
Claims (10)
1. An image processing method, characterized by comprising the steps of:
step S1, receiving an original image (1), and reducing the size of the original image (1) to obtain a first transition image (2);
step S2, providing a first expansion convolution neural network (30), and performing expansion convolution processing on the first transition image (2) by using the first expansion convolution neural network (30) so as to enhance the image quality of the first transition image and obtain a second transition image (3);
step S3, enlarging the size of the second transition image (3) to obtain a third transition image (4), wherein the size of the third transition image (4) is the same as that of the original image (1);
step S4, splicing the original image (1) and the third transition image (4) in a matrix mode to obtain a fourth transition image (5);
and step S5, providing a second expansion convolution neural network (60), and performing expansion convolution processing on the fourth transition image (5) by using the second expansion convolution neural network (60) to recover the detail information of the fourth transition image (5) to obtain the target image (6).
2. An image processing method as claimed in claim 1, characterized in that the size of the first transition image (2) is smaller than or equal to 256 pixels by 256 pixels.
3. An image processing method as claimed in claim 1, characterized in that the first dilated convolution neural network (30) performs an at least five-layer dilated convolution processing on the first transition image (2).
4. The image processing method of claim 1, characterized in that the second dilated convolution neural network (60) performs at least a two-layer dilated convolution processing on the fourth transition image (5).
5. The image processing method according to claim 1, wherein the size of the original image (1) is reduced by a bilinear interpolation algorithm in step S1, and the size of the second transition image (3) is enlarged by a bilinear interpolation algorithm in step S3.
6. An image processing system, comprising an input unit (10), a reducing unit (20) connected to the input unit (10), a first convolutional neural network (30) connected to the reducing unit (20), an amplifying unit (40) connected to the first convolutional neural network (30), a splicing unit (50) connected to the amplifying unit (40) and the input unit (10), a second convolutional neural network (60) connected to the splicing unit (50), and an output unit (70) connected to the second convolutional neural network (60);
the input unit (10) is used for receiving the original image (1) and providing the original image (1) to the reducing unit (20) and the splicing unit (50);
the reducing unit (20) is used for reducing the size of the original image (1) to obtain a first transition image (2);
the first expansion convolution neural network (30) is used for performing expansion convolution processing on the first transition image (2) so as to enhance the image quality of the first transition image and obtain a second transition image (3);
the amplifying unit (40) is used for amplifying the size of the second transition image (3) to obtain a third transition image (4), and the size of the third transition image (4) is the same as that of the original image (1);
the splicing unit (50) is used for performing matrix splicing on the original image (1) and the third transition image (4) to obtain a fourth transition image (5);
the second expansion convolution neural network (60) is used for performing expansion convolution processing on the fourth transition image (5) so as to recover the detail information of the fourth transition image (5) and obtain a target image (6);
the output unit (70) is used for outputting a target image (6).
7. An image processing system as claimed in claim 6, characterized in that the size of the first transition image (2) is smaller than or equal to 256 pixels by 256 pixels.
8. The image processing system of claim 6, wherein the first dilated convolution neural network (30) performs at least five-layer dilated convolution processing on the first transition image (2).
9. The image processing system of claim 6, wherein the second dilated convolution neural network (60) performs at least a two-layer dilated convolution processing on the fourth transition image (5).
10. The image processing system according to claim 6, wherein the reducing unit (20) reduces the size of the original image (1) by a bilinear interpolation algorithm, and the enlarging unit (40) enlarges the size of the second transition image (3) by the bilinear interpolation algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610059.5A CN110415187B (en) | 2019-07-04 | 2019-07-04 | Image processing method and image processing system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910610059.5A CN110415187B (en) | 2019-07-04 | 2019-07-04 | Image processing method and image processing system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110415187A CN110415187A (en) | 2019-11-05 |
| CN110415187B true CN110415187B (en) | 2021-07-23 |
Family
ID=68360596
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910610059.5A Active CN110415187B (en) | 2019-07-04 | 2019-07-04 | Image processing method and image processing system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415187B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021179147A1 (en) * | 2020-03-09 | 2021-09-16 | 华为技术有限公司 | Image processing method and apparatus based on neural network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108038893A (en) * | 2017-12-06 | 2018-05-15 | 深圳市未来媒体技术研究院 | A kind of generation method of 1,000,000,000 pixel videos based on Hybrid camera array |
| CN108376387A (en) * | 2018-01-04 | 2018-08-07 | 复旦大学 | Image deblurring method based on polymerization expansion convolutional network |
| US10068557B1 (en) * | 2017-08-23 | 2018-09-04 | Google Llc | Generating music with deep neural networks |
| CN108734211A (en) * | 2018-05-17 | 2018-11-02 | 腾讯科技（深圳）有限公司 | The method and apparatus of image procossing |
| CN108830809A (en) * | 2018-06-05 | 2018-11-16 | 陕西师范大学 | One kind is based on expansion convolved image denoising method |
| CN108876805A (en) * | 2018-06-20 | 2018-11-23 | 长安大学 | The end-to-end unsupervised scene of one kind can traffic areas cognition and understanding method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11587304B2 (en) * | 2017-03-10 | 2023-02-21 | Tusimple, Inc. | System and method for occluding contour detection |
-
2019
- 2019-07-04 CN CN201910610059.5A patent/CN110415187B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10068557B1 (en) * | 2017-08-23 | 2018-09-04 | Google Llc | Generating music with deep neural networks |
| CN108038893A (en) * | 2017-12-06 | 2018-05-15 | 深圳市未来媒体技术研究院 | A kind of generation method of 1,000,000,000 pixel videos based on Hybrid camera array |
| CN108376387A (en) * | 2018-01-04 | 2018-08-07 | 复旦大学 | Image deblurring method based on polymerization expansion convolutional network |
| CN108734211A (en) * | 2018-05-17 | 2018-11-02 | 腾讯科技（深圳）有限公司 | The method and apparatus of image procossing |
| CN108830809A (en) * | 2018-06-05 | 2018-11-16 | 陕西师范大学 | One kind is based on expansion convolved image denoising method |
| CN108876805A (en) * | 2018-06-20 | 2018-11-23 | 长安大学 | The end-to-end unsupervised scene of one kind can traffic areas cognition and understanding method |
Non-Patent Citations (4)
| Title |
|---|
| Accurate Gastric Cancer Segmentation in Digital Pathology Images Using Deformable Convolution and Multi-Scale Embedding Networks;Muyi Sun等;《IEEE Access》;20190524;第7卷;第75530-75541页 * |
| Salient object detection via multi-scale neural network;Weiqian Lu等;《Proceedings of the 10th International Conference on Internet Multimedia Computing and Service》;20180831;第1-5页 * |
| 基于上下文以及多尺度信息融合的目标检测算法;吕培建 等;《计算机科学》;20190615;第46卷(第6期);第279-283页 * |
| 基于全卷积神经网络的非对称并行语义分割模型;李宝奇;《电子学报》;20190515;第47卷(第5期);第1058-1064页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110415187A (en) | 2019-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10885384B2 (en) | Local tone mapping to reduce bit depth of input images to high-level computer vision tasks | |
| US8248494B2 (en) | Image dynamic range compression method, apparatus, and digital camera | |
| US20050117799A1 (en) | Method and apparatus for transforming a high dynamic range image into a low dynamic range image | |
| WO2023010754A1 (en) | Image processing method and apparatus, terminal device, and storage medium | |
| US20110285913A1 (en) | Enhanced histogram equalization | |
| JP2019067078A (en) | Image processing method and image processing program | |
| US11538136B2 (en) | System and method to process images of a video stream | |
| CN112602088A (en) | Method, system and computer readable medium for improving quality of low light image | |
| CN110889809B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN111565260A (en) | Haze reduction method and system for image processing | |
| CN111612722A (en) | Low-illumination image processing method based on simplified Unet full-convolution neural network | |
| CN113939845A (en) | Method, system and computer readable medium for improving image color quality | |
| WO2020118902A1 (en) | Image processing method and image processing system | |
| CN113096021A (en) | Image processing method, device, equipment and storage medium | |
| US11880963B2 (en) | Apparatus and method for image processing | |
| CN114240749A (en) | Image processing method, image processing device, computer equipment and storage medium | |
| CN109643528B (en) | Information processing apparatus, information processing method, and program | |
| Eilertsen | The high dynamic range imaging pipeline | |
| CN113781318A (en) | Image color mapping method and device, terminal equipment and storage medium | |
| CN109274949A (en) | A kind of method of video image processing and its device, display equipment | |
| JP2014119997A (en) | Image processing apparatus and control method thereof | |
| US20130287299A1 (en) | Image processing apparatus | |
| CN110415187B (en) | Image processing method and image processing system | |
| CN115115516A (en) | Real-world video super-resolution algorithm based on Raw domain | |
| JP2021114180A (en) | Image processing device, image processing method, and program |
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 | ||
| CB02 | Change of applicant information |
Address after: 9-2 Tangming Avenue, Guangming New District, Shenzhen City, Guangdong Province Applicant after: TCL Huaxing Photoelectric Technology Co.,Ltd. Address before: 9-2 Tangming Avenue, Guangming New District, Shenzhen City, Guangdong Province Applicant before: Shenzhen China Star Optoelectronics Technology Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |