CN112651911B - High dynamic range imaging generation method based on polarized image - Google Patents
High dynamic range imaging generation method based on polarized image Download PDFInfo
- Publication number
- CN112651911B CN112651911B CN202011388884.4A CN202011388884A CN112651911B CN 112651911 B CN112651911 B CN 112651911B CN 202011388884 A CN202011388884 A CN 202011388884A CN 112651911 B CN112651911 B CN 112651911B
- Authority
- CN
- China
- Prior art keywords
- polarization
- image
- hdr
- polarized
- images
- 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
- 238000003384 imaging method Methods 0.000 title claims abstract description 36
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000010287 polarization Effects 0.000 claims abstract description 100
- 230000004927 fusion Effects 0.000 claims abstract description 15
- 230000000694 effects Effects 0.000 claims abstract description 14
- 238000010586 diagram Methods 0.000 claims abstract description 13
- 238000004590 computer program Methods 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000013459 approach Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 230000004313 glare Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 230000002238 attenuated effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/70—Circuitry for compensating brightness variation in the scene
- H04N23/741—Circuitry for compensating brightness variation in the scene by increasing the dynamic range of the image compared to the dynamic range of the electronic image sensors
-
- 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
- G06T2207/20081—Training; Learning
-
- 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]
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Studio Devices (AREA)
Abstract
The application discloses a high dynamic range imaging generation method based on polarized images, which comprises the following steps: aiming at a target area to be imaged, acquiring a plurality of polarized images with different polarization angles through single exposure of a polarization camera; constructing an HDR generation model, using a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model; calculating linear polarization degree characteristics and polarization angle characteristics by using the polarization image; and taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic graphs of all the polarized images output by the HDR generation model into an HDR image. According to the method, the HDR image is obtained through network learning and using polarization degree characteristics and the like, a plurality of images with different exposure time are not needed, the HDR image can be obtained through single exposure, the problem of image double image generated by fusion of a plurality of images is effectively reduced, and the imaging effect is improved.
Description
Technical Field
The application relates to the technical field of high dynamic range imaging, in particular to a high dynamic range imaging generation method based on polarized images.
Background
High dynamic range imaging (High Dynamic Range Imaging, HDR) is a group of techniques used to achieve a larger dynamic range of exposure than conventional digital image techniques, and is widely used in the fields of computer graphics, image photography, and the like. The common digital image has 256 brightness levels, which is far inferior to the perception range of human eyes, so that the HDR image can provide more dynamic range and image details, so that the HDR image better reflects the visual effect in the real environment of human. Traditional HDR imaging mainly comprises the steps of shooting a plurality of images with different exposure time, and synthesizing a final HDR image by utilizing the optimal detail corresponding to each exposure time; however, the existing HDR imaging method is easy to generate image ghosting problem in the process of fusing multiple images with different exposure time, and the imaging effect needs to be improved.
Disclosure of Invention
The application aims to provide a high dynamic range imaging generation method based on polarized images, which is used for solving the problems of easy generation of double images and poor imaging effect existing in the existing HDR imaging.
In order to realize the tasks, the application adopts the following technical scheme:
in a first aspect, the present application provides a method for generating high dynamic range imaging based on polarized images, comprising:
aiming at a target area to be imaged, acquiring a plurality of polarized images with different polarization angles through single exposure of a polarization camera;
constructing an HDR generation model, using a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model;
calculating linear polarization degree characteristics and polarization angle characteristics by using the polarization image;
and taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic graphs of all polarized images output by the HDR generation model into an HDR image.
Further, the polarization images with different polarization angles are polarization images with 4 polarization angles in total of 0 °, 45 °, 90 ° and 135 °.
Further, the calculation formulas of the polarization degree characteristic P1 and the polarization angle characteristic P2 are as follows:
in the above, I 0 ,I 45 ,I 90 ,I 135 The intensities of the polarized images obtained at the polarization angles of 0 °, 45 °, 90 °, 135 °, respectively, are shown.
Further, the HDR generation model is constructed by utilizing a convolutional neural network; the HDR generation model adopts a structure of an encoder-decoder, a downsampling structure of the encoder is used for extracting the characteristics of the polarized image, and the decoder restores the characteristics of the image to the original image size; in the model, granularity coarseness of upsampling is improved by introducing jump connections from the high resolution feature map.
Further, a loss function of the model is generated by constructing the HDR, so that the model output approaches to the true value of the HDR image; the loss function is expressed as:
in the above, mu x 、Mean value, standard deviation, mu of characteristic diagram representing polarized image x output by model y 、/>Mean value, standard deviation, sigma representing true value of image corresponding to polarized image x xy Covariance representing true values of the feature map, image, C 1 、C 2 Is constant.
In a second aspect, the present application provides a polarized image-based high dynamic range imaging generating apparatus, comprising:
the acquisition module is used for acquiring a plurality of polarized images with different polarization angles through single exposure of the polarization camera aiming at a target area needing imaging;
the model construction module is used for constructing an HDR generation model, taking a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model;
the feature calculation module is used for calculating linear polarization degree features and polarization angle features by using the polarized image;
and the image fusion module is used for taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic images of all the polarized images output by the HDR generation model into an HDR image.
In a third aspect, the present application provides a terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the polarization image based high dynamic range imaging generation method of the first aspect when executing the computer program.
In a fourth aspect, the present application provides a computer readable storage medium storing a computer program which when executed by a processor implements the steps of the polarization image based high dynamic range imaging generation method of the first aspect described above.
Compared with the prior art, the application has the following technical characteristics:
1. different from the scheme that the existing high dynamic range imaging technology needs to fuse a plurality of images with different exposure time, the method adopts different polarized images captured by single exposure of a polarized camera as network input, obtains an HDR image through network learning, using polarization degree characteristics and the like, does not need a plurality of images with different exposure time, can obtain the HDR image through single exposure, and effectively reduces the problems of image ghosts and the like generated by fusion of a plurality of images.
2. Compared with the existing HDR method for generating a single image based on a deep learning model, the method fully utilizes the characteristic that a polarization camera can filter the glare on the surface of an object, restores the original image information on the surface of the object, improves the imaging effect, and is more advantageous in HDR imaging.
Drawings
FIG. 1 is a schematic flow chart of the method of the present application;
FIG. 2 is a schematic view of capturing four images with different polarization angles according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an HDR generation model;
FIG. 4 is a schematic diagram of an HDR fusion flow;
FIG. 5 is a schematic diagram of a polarized camera filtering object surface glare and recovering object surface information according to an embodiment of the present application.
Detailed Description
The strong learning capability of the deep learning enables the deep learning to have more excellent performance than the traditional image processing method in the image field, so that the HDR image generated by combining the single exposure image with the deep learning has larger research potential and can overcome the defects of the traditional HDR image.
The application provides a deep learning model for synthesizing a high dynamic image by single exposure images, which is characterized in that parameters such as polarization degree characteristics, polarization angle characteristics and the like are calculated through single exposure polarized pictures captured by a polarization camera, and the high dynamic range imaging deep learning model is trained by combining the polarized images, so that an HDR image can be finally generated.
Referring to fig. 1, the high dynamic range imaging generation method based on polarized image of the present application comprises the following steps:
step 1, aiming at a target area needing imaging, acquiring a plurality of polarized images with different polarization angles through single exposure of a polarization camera.
The polarization camera can capture four images with different polarization angles at a time, approximately view the images as images with different exposure times, and the polaroid can effectively restore the original surface information of the object in the scene.
Due to the polarization effect of light, the magnitudes of the corresponding pixels of the four polarized images captured by the polarized camera from the non-uniformly polarized scene are not the same, because the different angles of the polarizers on the four pixels cause the illumination intensity to be attenuated; as shown in fig. 2, the ground in the images is exposed differently in each image due to polarization effects.
Capturing polarized images using a polarized camera can be understood as imaging the same scene at different exposure times, so that there is a possibility to generate HDR images from four polarized images. Since the irradiance values are measured four times at four pixels and the filtering effect of the polarizer would be such that one of the four pixels is in a non-overexposed region, the HDR image would be significantly improved in the dark or underexposed region.
And 2, constructing an HDR generation model by adopting a convolutional neural network, taking a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic map of each polarized image by using the HDR generation model.
And 3, calculating linear polarization degree characteristics P1 and polarization angle characteristics P2 by using the polarized image.
And 4, fusing the feature images of all polarized images output by the HDR generation model into an HDR image by using P1 and P2 as fusion parameters and using a fusion algorithm.
In the embodiment of the application, a polarization camera obtains polarized images of 4 polarized angles of 0 °, 45 °, 90 ° and 135 ° in a single exposure, and the calculation formulas of the polarization degree characteristic P1 and the polarization angle characteristic P2 are as follows:
in the above, I 0 ,I 45 ,I 90 ,I 135 The intensities of the polarized images obtained at the polarization angles of 0 °, 45 °, 90 °, 135 °, respectively, are shown.
The HDR generation model adopts an encoder-decoder structure, as shown in fig. 3, where the downsampling structure of the encoder can extract the features of the polarized image, and the decoder restores the features of the image to the original image size. Since upsampling results in loss of image information, the present scheme introduces a jump connection from the high resolution feature map to improve the granularity roughness of the upsampling. Generating a loss function of the model by constructing the HDR, so that the model output approaches to the true value of the HDR image; the loss function in this scheme is constructed as follows:
in the above, mu x 、Mean value, standard deviation, mu of characteristic diagram representing polarized image x output by model y 、/>Mean value, standard deviation, sigma representing true value of image corresponding to polarized image x xy Covariance representing true values of the feature map, image, C 1 、C 2 Is constant.
As shown in fig. 4, after obtaining a feature map output by the HDR generation model for each polarized image, the feature map, the polarization degree feature P1, and the polarization angle feature P2 are combined and fused, so as to obtain a final HDR image.
The scheme fully utilizes the characteristic that the polarization camera can filter the glare of the object surface, restores the original image information of the object surface, and restores the lost information of the overexposed area, as shown in fig. 5.
According to another aspect of the present application, there is provided a high dynamic range imaging generation apparatus based on a polarized image, comprising:
the acquisition module is used for acquiring a plurality of polarized images with different polarization angles through single exposure of the polarization camera aiming at a target area needing imaging;
the model construction module is used for constructing an HDR generation model, taking a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model;
the feature calculation module is used for calculating linear polarization degree features and polarization angle features by using the polarized image;
and the image fusion module is used for taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic images of all the polarized images output by the HDR generation model into an HDR image.
The specific execution steps of the above modules are the same as the corresponding steps in the foregoing method embodiments, and are not described herein.
The embodiment of the application further provides terminal equipment, which can be a computer or a server; comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the above-described polarization image based high dynamic range imaging generation method when the computer program is executed.
A computer program may also be split into one or more modules/units that are stored in a memory and executed by a processor to perform the present application. One or more modules/units may be a series of instruction segments of a computer program capable of performing a specific function, where the instruction segments are used to describe an execution process of the computer program in the terminal device, for example, the computer program may be divided into an acquisition module, a model building module, a feature calculation module and an image fusion module, and the functions of the modules are referred to in the foregoing apparatus and are not described herein.
The implementation of the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the above-described polarization image-based high dynamic range imaging generation method.
The integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present application may implement all or part of the flow in the method of the above embodiment, and may also be implemented by a computer program to instruct related hardware. The computer readable medium may include: any entity or device capable of carrying computer program code, a recording medium, a USB flash disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory ROM, a random access memory RAM, an electrical carrier wave signal, a telecommunication signal, a software distribution medium, and so forth.
The above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.
Claims (5)
1. A high dynamic range imaging generation method based on polarized images, comprising:
aiming at a target area to be imaged, acquiring a plurality of polarized images with different polarization angles through single exposure of a polarization camera;
constructing an HDR generation model, using a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model;
calculating linear polarization degree characteristics and polarization angle characteristics by using the polarization image;
taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic images of all polarized images output by the HDR generation model into an HDR image;
the polarization images with different polarization angles are polarization images with 4 polarization angles in total, namely 0 degree, 45 degrees, 90 degrees and 135 degrees;
the calculation formulas of the polarization degree characteristic P1 and the polarization angle characteristic P2 are as follows:
in the above, I 0 ,I 45 ,I 90 ,I 135 Respectively representing the intensities of polarized images acquired at polarized angles of 0 °, 45 °, 90 °, 135 °;
generating a loss function of the model by constructing the HDR, so that the model output approaches to the true value of the HDR image; the loss function is expressed as:
in the above, mu x 、Mean value, standard deviation, mu of characteristic diagram representing polarized image x output by model y 、/>Mean value, standard deviation, sigma representing true value of image corresponding to polarized image x xy Covariance representing true values of the feature map, image, C 1 、C 2 Is constant.
2. The polarized image based high dynamic range imaging generation method of claim 1, wherein the HDR generation model is constructed using a convolutional neural network; the HDR generation model adopts a structure of an encoder-decoder, a downsampling structure of the encoder is used for extracting the characteristics of the polarized image, and the decoder restores the characteristics of the image to the original image size; in the model, granularity coarseness of upsampling is improved by introducing jump connections from the high resolution feature map.
3. A polarized image-based high dynamic range imaging generation apparatus, comprising:
the acquisition module is used for acquiring a plurality of polarized images with different polarization angles through single exposure of the polarization camera aiming at a target area needing imaging;
the model construction module is used for constructing an HDR generation model, taking a plurality of polarized images with different polarization angles as an input of the HDR generation model, learning the characteristics of the polarized images and the characteristics of polarization effects by using the HDR generation model, and finally outputting a characteristic diagram of each polarized image by using the HDR generation model;
the feature calculation module is used for calculating linear polarization degree features and polarization angle features by using the polarized image;
the image fusion module is used for taking the polarization degree characteristics and the polarization angle characteristics as fusion parameters, and fusing the characteristic images of all polarized images output by the HDR generation model into an HDR image;
the polarization images with different polarization angles are polarization images with 4 polarization angles in total, namely 0 degree, 45 degrees, 90 degrees and 135 degrees;
the calculation formulas of the polarization degree characteristic P1 and the polarization angle characteristic P2 are as follows:
in the above, I 0 ,I 45 ,I 90 ,I 135 Respectively representing the intensities of polarized images acquired at polarized angles of 0 °, 45 °, 90 °, 135 °;
generating a loss function of the model by constructing the HDR, so that the model output approaches to the true value of the HDR image; the loss function is expressed as:
in the above, mu x 、Mean value, standard deviation, mu of characteristic diagram representing polarized image x output by model y 、/>Mean value, standard deviation, sigma representing true value of image corresponding to polarized image x xy Covariance representing true values of the feature map, image, C 1 、C 2 Is constant.
4. Terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the polarization image based high dynamic range imaging generation method according to any of claims 1 to 2 when the computer program is executed.
5. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the polarization image based high dynamic range imaging generation method according to any one of claims 1 to 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011388884.4A CN112651911B (en) | 2020-12-01 | 2020-12-01 | High dynamic range imaging generation method based on polarized image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011388884.4A CN112651911B (en) | 2020-12-01 | 2020-12-01 | High dynamic range imaging generation method based on polarized image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112651911A CN112651911A (en) | 2021-04-13 |
| CN112651911B true CN112651911B (en) | 2023-10-13 |
Family
ID=75350062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011388884.4A Active CN112651911B (en) | 2020-12-01 | 2020-12-01 | High dynamic range imaging generation method based on polarized image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112651911B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113487504B (en) * | 2021-07-01 | 2022-12-09 | 天津大学 | Underwater gray level polarization image restoration method based on generation countermeasure network |
| CN113962915B (en) * | 2021-10-20 | 2022-04-29 | 哈尔滨工业大学 | Self-adaptive nonlinear super-dynamic image synthesis method under non-uniform illumination condition |
| CN115376009B (en) * | 2022-06-09 | 2023-04-18 | 中国矿业大学(北京) | Shadow vegetation extraction method and system based on polarization image |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657965A (en) * | 2015-03-12 | 2015-05-27 | 长春理工大学 | Polarized image fusion method based on discrete continuous curvelet |
| EP2889807A2 (en) * | 2013-12-24 | 2015-07-01 | Ricoh Company, Ltd. | Image processing device, image processing system, mobile object control apparatus, image processing method, and carrier means |
| CN105139367A (en) * | 2015-07-27 | 2015-12-09 | 中国科学院光电技术研究所 | Visible light polarization image fusion method based on non-subsampled shear wave |
| CN111669514A (en) * | 2020-06-08 | 2020-09-15 | 北京大学 | High dynamic range imaging method and apparatus |
| CN111986129A (en) * | 2020-06-30 | 2020-11-24 | 普联技术有限公司 | HDR image generation method and device based on multi-shot image fusion and storage medium |
-
2020
- 2020-12-01 CN CN202011388884.4A patent/CN112651911B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2889807A2 (en) * | 2013-12-24 | 2015-07-01 | Ricoh Company, Ltd. | Image processing device, image processing system, mobile object control apparatus, image processing method, and carrier means |
| CN104657965A (en) * | 2015-03-12 | 2015-05-27 | 长春理工大学 | Polarized image fusion method based on discrete continuous curvelet |
| CN105139367A (en) * | 2015-07-27 | 2015-12-09 | 中国科学院光电技术研究所 | Visible light polarization image fusion method based on non-subsampled shear wave |
| CN111669514A (en) * | 2020-06-08 | 2020-09-15 | 北京大学 | High dynamic range imaging method and apparatus |
| CN111986129A (en) * | 2020-06-30 | 2020-11-24 | 普联技术有限公司 | HDR image generation method and device based on multi-shot image fusion and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于多尺度边缘表示的偏振图像二次融合算法;李世维;黄丹飞;刘辉;王惠敏;;激光与红外(01);第115-120页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112651911A (en) | 2021-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110428366B (en) | Image processing method and device, electronic equipment and computer readable storage medium | |
| CN112651911B (en) | High dynamic range imaging generation method based on polarized image | |
| JP7003238B2 (en) | Image processing methods, devices, and devices | |
| Jiang et al. | Learning to see moving objects in the dark | |
| Cui et al. | Multitask aet with orthogonal tangent regularity for dark object detection | |
| Lee et al. | Deep chain hdri: Reconstructing a high dynamic range image from a single low dynamic range image | |
| WO2020259118A1 (en) | Method and device for image processing, method and device for training object detection model | |
| WO2021179820A1 (en) | Image processing method and apparatus, storage medium and electronic device | |
| WO2020152521A1 (en) | Systems and methods for transforming raw sensor data captured in low-light conditions to well-exposed images using neural network architectures | |
| WO2021057474A1 (en) | Method and apparatus for focusing on subject, and electronic device, and storage medium | |
| CN113168670A (en) | Bright spot removal using neural networks | |
| US9473700B2 (en) | Camera systems and methods for gigapixel computational imaging | |
| CN111986084A (en) | Multi-camera low-illumination image quality enhancement method based on multi-task fusion | |
| CN108769523B (en) | Image processing method and device, electronic equipment and computer readable storage medium | |
| CN113658065B (en) | Image noise reduction method and device, computer readable medium and electronic equipment | |
| Aakerberg et al. | RELLISUR: A real low-light image super-resolution dataset | |
| An et al. | Single-shot high dynamic range imaging via deep convolutional neural network | |
| KR20220028962A (en) | Image enhancement method, image enhancement apparatus, method and apparatus of traning thereof | |
| CN111986106A (en) | High dynamic image reconstruction method based on neural network | |
| Zheng et al. | Neural augmented exposure interpolation for two large-exposure-ratio images | |
| Zhang et al. | Deep motion blur removal using noisy/blurry image pairs | |
| CN110827375B (en) | Infrared image true color coloring method and system based on low-light-level image | |
| CN108122218B (en) | Image fusion method and device based on color space | |
| CN109325912B (en) | Reflection separation method based on polarized light field and calibration splicing system | |
| JP5843599B2 (en) | Image processing apparatus, imaging apparatus, and method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |