CN114299076B - Depth image cavity filling method and device based on discrete wavelet decomposition - Google Patents
Depth image cavity filling method and device based on discrete wavelet decomposition Download PDFInfo
- Publication number
- CN114299076B CN114299076B CN202111329375.9A CN202111329375A CN114299076B CN 114299076 B CN114299076 B CN 114299076B CN 202111329375 A CN202111329375 A CN 202111329375A CN 114299076 B CN114299076 B CN 114299076B
- Authority
- CN
- China
- Prior art keywords
- image
- frequency information
- information image
- filling
- low
- 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
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a depth image cavity filling method and device based on discrete wavelet decomposition, wherein the method comprises the following steps: acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image; filling the cavity points in the first low-frequency information image; performing binarization processing on the original depth image, and performing discrete wavelet transformation on the obtained binarized depth image to obtain a second low-frequency information image and a second high-frequency information image; repairing the first high-frequency information image by using the second high-frequency information image; performing inverse discrete wavelet transform on the first low-frequency information image after filling the cavity and the first high-frequency information image after repairing to obtain an initial image after filling the cavity; and carrying out fusion processing on the initial image filled with the cavity and the original depth image to obtain a final depth image. The method provided by the invention improves the cavity filling efficiency and generalization capability of the depth image and improves the filling effect.
Description
Technical Field
The invention belongs to the technical field of depth image processing, and particularly relates to a depth image cavity filling method and device based on discrete wavelet decomposition.
Background
With the development of computer vision technology, depth images play an important role in applications such as object recognition and three-dimensional reconstruction. Compared with the traditional gray level image and color image, the depth image has three-dimensional characteristic information of the object, and can reflect the depth information of the shot object, so that the method is increasingly applied to the fields of computer vision, computer graphics and the like.
Currently, depth cameras are one of the main approaches to acquire depth images. The Kinect depth camera has the advantages of low cost, capability of simultaneously acquiring depth images and RGB images and the like, and is widely applied to the fields of intelligent robots such as environment sensing, target identification and positioning, semantic segmentation and the like. However, due to factors such as occlusion and measurement range limitation, a hole exists in the depth image acquired by the Kinect depth camera. For this reason, hole filling is required for the depth image to improve the subsequent depth image processing quality.
The existing depth image cavity filling methods mainly comprise two types: one is a depth image hole filling method based on a joint bilateral filter (Joint bilateral filtering, JBF). The method can utilize the matched gray level image and the depth image to carry out hole filling on the depth image, but is difficult to carry out hole filling on the depth image directly, and the method has higher calculation complexity. The other type is a depth image cavity region filling method based on a neural network. The hole filling effect of the method is greatly influenced by the training data set, and is good for scenes similar to the training data set, but poor for depth images with larger differences from the training data set.
Thus, the existing depth image cavity filling method mainly has two problems: firstly, the algorithm is complex, and the filling efficiency is low; secondly, for the depth image with larger difference with the training data set, the hole filling effect is poor, and the generalization capability is insufficient.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a depth image cavity filling method and device based on discrete wavelet decomposition. The technical problems to be solved by the invention are realized by the following technical scheme:
in a first aspect, the present invention provides a depth image hole filling method based on discrete wavelet decomposition, including:
acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image;
filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image with the filled cavities;
performing binarization processing on the original depth image to obtain a binarized depth image;
performing discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image;
repairing the first high-frequency information image by using the second high-frequency information image to obtain a repaired first high-frequency information image;
performing discrete wavelet inverse transformation on the first low-frequency information image after cavity filling and the first high-frequency information image after repairing to obtain an initial image after cavity filling;
and carrying out fusion processing on the original depth image and the initial image filled with the cavity to obtain a final depth image.
In one embodiment of the present invention, filling the hole point in the first low frequency information image to obtain a first low frequency information image after filling the hole, including:
obtaining a hole point and a non-hole point in the image according to the depth value of each pixel point in the first low-frequency information image;
calculating the similarity weight of the current hole point and each non-hole point in the neighborhood of the current hole point;
calculating a depth estimation value of the current cavity point according to the similarity weight, and filling the current cavity point with the depth estimation value;
traversing the first low-frequency information image to fill all the cavity points, and obtaining the first low-frequency information image after cavity filling.
In one embodiment of the present invention, performing binarization processing on the original depth image to obtain a binarized depth image, including:
acquiring a hole point and a non-hole point in the original depth image;
and performing binarization processing on the original depth image according to the mode that the hole point is 0 and the non-hole point is 1 to obtain a binarized depth image.
In one embodiment of the present invention, repairing the first high frequency information image by using the second high frequency information image to obtain a repaired first high frequency information image includes:
comparing the first high-frequency information image with the second high-frequency information image, and setting zero to the pixel value of the corresponding position in the first high-frequency information image when the pixel value of the same position is not zero so as to obtain a repaired first high-frequency information image.
In one embodiment of the present invention, the fusing processing of the initial image after the hole filling and the original depth image includes:
traversing the initial image filled with the holes, and carrying out depth filling again on each pixel point, if the corresponding point of the current point in the original depth image is a non-hole point, replacing the depth value of the current point by the depth value of the point in the original depth image; otherwise, the depth value of the current point is reserved.
In a second aspect, the present invention provides a depth image cavity filling device based on discrete wavelet decomposition, including:
the first transformation module is used for acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image;
the low-frequency filling module is used for filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image with the cavity filled;
the binarization module is used for carrying out binarization processing on the original depth image to obtain a binarized depth image;
the second transformation module is used for carrying out discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image;
the high-frequency restoration module is used for restoring the first high-frequency information image by utilizing the second high-frequency information image to obtain a restored first high-frequency information image;
the inverse transformation module is used for performing discrete wavelet inverse transformation on the first low-frequency information image after cavity filling and the first high-frequency information image after repairing to obtain an initial image after cavity filling;
and the fusion processing module is used for carrying out fusion processing on the initial image filled with the cavity and the original depth image to obtain a final depth image.
In a third aspect, the invention provides a depth image cavity filling electronic device based on discrete wavelet decomposition, which comprises a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface, and the memory are communicated with each other through the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the method steps in the embodiment when executing the program stored in the memory.
In a fourth aspect, the present invention provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the computer program realizes the method steps described in the above embodiments.
The invention has the beneficial effects that:
according to the invention, the discrete wavelet transformation is carried out on the original depth image to obtain the high-frequency and low-frequency decomposition image, the cavity point detection and filling are carried out on the low-frequency information part, the high-frequency information part is subjected to contrast restoration, and finally the fusion treatment is carried out on the obtained cavity filled image by utilizing the original depth image, so that the cavity filling efficiency and generalization capability of the depth image are improved, and meanwhile, the filling effect is improved.
The present invention will be described in further detail with reference to the accompanying drawings and examples.
Drawings
Fig. 1 is a schematic flow chart of a depth image cavity filling method based on discrete wavelet decomposition according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a depth image cavity filling method based on discrete wavelet decomposition according to an embodiment of the present invention;
FIG. 3 is a comparison of an original depth image and a depth image after filling of a cavity according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a depth image cavity filling device based on discrete wavelet decomposition according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a depth image cavity filling electronic device based on discrete wavelet decomposition according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a computer-readable storage medium according to an embodiment.
Detailed Description
The present invention will be described in further detail with reference to specific examples, but embodiments of the present invention are not limited thereto.
Example 1
Referring to fig. 1, fig. 1 is a flow chart of a depth image cavity filling method based on discrete wavelet decomposition according to an embodiment of the present invention, including the following steps:
step 1: and obtaining an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image.
First, an original depth image is acquired. In this embodiment, the original depth image is denoted as U 0 . Specifically, the method for acquiring the original depth image is not limited, and the original depth image may be acquired by any one of the existing methods, such as a laser radar depth imaging method, a computer stereoscopic imaging method, a coordinate measuring machine method, a moire fringe method, a structured light method, and the like.
Then, for the acquired original depth image U 0 Performing discrete wavelet transform to divide itThe resolution is divided into a low frequency part and a high frequency part, and the low frequency part and the high frequency part are recorded as a first low frequency information image and a first high frequency information image. Wherein the first low frequency information image is denoted cA; the first high-frequency information image comprises three images which are respectively marked as cH, cV and cD, wherein cH is high-frequency decomposition information for describing the transverse details of the original image; cV is high-frequency decomposition information for describing the vertical detail of the original image; cD is high frequency decomposition information that characterizes the detail on the diagonal of the original image.
It should be noted that, the specific implementation manner of the discrete wavelet transform algorithm in this embodiment is a mature technology in the prior art, and will not be described in detail here.
Step 2: and filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image with the filled cavities.
Specifically, in this embodiment, for each hole point x in the first low-frequency information image cA, each non-hole point y e I (x) in the neighborhood is used to perform hole filling on each hole point x, so as to obtain a hole-filled image. The method mainly comprises the following substeps:
21 And obtaining the hole point and the non-hole point in the image according to the depth value of each pixel point in the first low-frequency information image.
Specifically, in this embodiment, whether a certain pixel point in the first low-frequency information image cA is a hole point is determined according to the depth value of each pixel point, when the depth value of the certain pixel point is zero, the pixel point is considered to be a hole point, and correspondingly, a point with a depth value not zero is a non-hole point.
22 Calculating the similarity weight of the current hole point and each non-hole point in the neighborhood of the current hole point.
Considering that the depth values of adjacent pixel points in the depth image are relatively close, the closer the distance between the non-hole point y and the hole point x is, the larger the influence of the point y on the depth value of the hole point x is considered. Therefore, the embodiment is convenient to calculate the similarity weight of the hole point and each non-hole point in the adjacent area by setting a distance weight parameter.
First, for a certain hole point x in the first low-frequency information image, a distance weight d (x, y) between x and each non-hole point y e I (x) in the neighborhood is defined as follows:
wherein sigma is a control parameter of the decay rate of the exponential function.
Then, according to the distance weight d (x, y), a similarity weight ω (x, y) of the non-hole point y and the hole point x is defined as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,is a normalization constant.
23 Calculating the depth estimation value of the current hole point according to the similarity weight, and filling the current hole point with the depth estimation value.
Specifically, according to the depth value u (y) and the similarity weight ω (x, y) of each non-hole point y e I (x) in the vicinity of the hole point x, estimating the depth value of the hole point x by using a weighted average method, and then the depth estimation value after filling the hole point x is:
and filling the current hole point x according to the obtained depth estimation value.
24 Traversing the first low-frequency information image to fill all the hole points to obtain the first low-frequency information image filled with the holes.
Filling all the cavity points in the first low-frequency information image cA by adopting the method so as to obtain a first low-frequency information image after cavity filling, and marking asAn image.
Step 3: and carrying out binarization processing on the original depth image to obtain a binarized depth image.
Specifically, in this embodiment, the original depth image U is determined according to whether the depth image U is a hole point 0 And (5) performing binarization processing. Firstly, acquiring hole points and non-hole points in the original depth image; then, binarizing the original depth image according to the mode that the hole point is 0 and the non-hole point is 1 to obtain a binarized depth image U 1 。
Step 4: and performing discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image.
Referring to fig. 2, fig. 2 is a schematic diagram of a depth image cavity filling method based on discrete wavelet decomposition according to an embodiment of the present invention.
Specifically, referring to the above step 1, the binarized depth image U may be obtained 1 Discrete wavelet transform is performed to obtain a second low-frequency information image cA1 and second high-frequency information images cH1, cV1, cD1, respectively.
Step 5: and repairing the first high-frequency information image by using the second high-frequency information image to obtain a repaired first high-frequency information image.
First, a cH image in a first high-frequency information image (i.e., a high-frequency information portion of an original depth image) and a second high-frequency information image (i.e., a binarized depth image U 1 Is used for comparing the cH1 image in the high-frequency information part) and setting the pixel value of the corresponding position z in the cH to zero when the pixel value of the same position z is not zero so as to obtain the repaired high-frequency information part
Likewise, the cV image and the cD image are compared with the cV1 image and the cD1 image, respectively, to obtain a repaired high-frequency information portionAnd->
Thus, the repaired first high-frequency information image is obtained
Step 6: and performing inverse discrete wavelet transformation on the first low-frequency information image after filling the cavity and the first high-frequency information image after repairing to obtain an initial image after filling the cavity.
Specifically, for the first low-frequency information image after filling the cavity obtained in step 2And the repaired first high-frequency information image obtained in the step 5->Performing inverse discrete wavelet transform to obtain initial depth image U after cavity filling 2 。
Step 7: and carrying out fusion processing on the initial image filled with the cavity and the original depth image to obtain a final depth image.
Due to the specificity of discrete wavelet transform, the resulting image U 2 With noise, in order to eliminate the effect of noise, the present embodiment obtains an initial image U after filling the cavity 2 After that, the original depth image U is utilized 0 Refilling the depth value to obtain a final depth image after cavity filling
Specifically, the initial image U after the cavity filling is traversed 2 And re-filling each pixel point, if the current point x is in the original depth image U 0 If the corresponding point in the original depth image U is a non-hole point, the original depth image U is utilized 0 The depth value of the point replaces the depth value of the current point; otherwise, the depth value of the current point is reserved.
That is, the final depth imageThe method meets the following conditions:
that is, in the final fill image, the current point is U 0 Depth value of non-hole point of (a) is image U 0 Depth value of (1), current point is U 0 The depth value of the cavity point is U 2 Depth value of the image.
According to the embodiment, the original depth image is subjected to discrete wavelet transformation to obtain the high-frequency and low-frequency decomposition image, the low-frequency information part is subjected to detection and filling of the cavity points, the high-frequency information part is subjected to contrast restoration, and finally the obtained cavity filled image is subjected to fusion processing by utilizing the original depth image, so that the cavity filling efficiency and generalization capability of the depth image are improved, and meanwhile, the filling effect is improved.
In order to verify the beneficial effects of the invention, four depth images with cavity points in different scenes are also selected, and are processed according to the method which is powerful, so that the processed images are obtained.
Referring to fig. 3, fig. 3 is a comparison chart of an original depth image and a depth image after filling a hole, where a first row (a) is an original depth image with hole points of different scenes, and a second row (b) is a corresponding depth image after filling a hole.
As is apparent from fig. 3, the depth image cavity filling method based on discrete wavelet decomposition provided in this embodiment has a good filling effect.
Example two
On the basis of the first embodiment, the present embodiment provides a depth image hole filling device based on discrete wavelet decomposition. Referring to fig. 4, fig. 4 is a schematic structural diagram of a depth image cavity filling device based on discrete wavelet decomposition according to an embodiment of the present invention, which includes:
the first transformation module 1 is used for acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image;
the low-frequency filling module 2 is used for filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image after cavity filling;
the binarization module 3 is used for performing binarization processing on the original depth image to obtain a binarized depth image;
a second transformation module 4, configured to perform discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image;
the high-frequency restoration module 5 is used for restoring the first high-frequency information image by using the second high-frequency information image to obtain a restored first high-frequency information image;
the inverse transformation module 6 is used for performing inverse discrete wavelet transformation on the first low-frequency information image after cavity filling and the first high-frequency information image after repairing to obtain an initial image after cavity filling;
and the fusion processing module 7 is used for carrying out fusion processing on the initial image filled with the cavity and the original depth image to obtain a final depth image.
The depth image cavity filling device based on discrete wavelet decomposition provided in this embodiment may be used to implement the method provided in the first embodiment, and detailed processes are not described herein.
Therefore, the depth image cavity filling device based on discrete wavelet decomposition can improve the depth image cavity filling efficiency and generalization capability and improve the filling effect.
Example III
On the basis of the first embodiment, the present embodiment provides a depth image hole filling electronic device based on discrete wavelet decomposition, referring to fig. 5, fig. 5 is a schematic structural diagram of the depth image hole filling electronic device based on discrete wavelet decomposition, which includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;
a memory for storing a computer program;
the processor is configured to implement the method steps provided in the first embodiment when executing the program stored in the memory, and the implementation principle and the technical effect are similar, and are not described herein again.
Example IV
On the basis of the above embodiment, please refer to fig. 6, fig. 6 is a schematic diagram illustrating a structure of a computer readable storage medium according to an embodiment of the present invention. The computer readable storage medium provided in this embodiment stores a computer program, which when executed by a processor, implements the method steps provided in the first embodiment, and the implementation principle and technical effects are similar, and are not described herein again.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (7)
1. A depth image cavity filling method based on discrete wavelet decomposition is characterized by comprising the following steps:
acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image;
filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image with the filled cavities;
performing binarization processing on the original depth image to obtain a binarized depth image;
performing discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image;
performing contrast restoration on the first high-frequency information image by using the second high-frequency information image to obtain a restored first high-frequency information image;
performing discrete wavelet inverse transformation on the first low-frequency information image after cavity filling and the first high-frequency information image after repairing to obtain an initial image after cavity filling;
fusing the initial image filled with the cavity with the original depth image to obtain a final depth image;
filling the hole points in the first low-frequency information image to obtain a first low-frequency information image filled with holes, wherein the step of obtaining the first low-frequency information image filled with holes comprises the following steps:
obtaining a hole point and a non-hole point in the image according to the depth value of each pixel point in the first low-frequency information image;
calculating the similarity weight of the current hole point and each non-hole point in the neighborhood of the current hole point;
calculating a depth estimation value of the current cavity point according to the similarity weight, and filling the current cavity point with the depth estimation value;
traversing the first low-frequency information image to fill all the cavity points, and obtaining the first low-frequency information image after cavity filling.
2. The depth image hole filling method based on discrete wavelet decomposition according to claim 1, wherein performing binarization processing on the original depth image to obtain a binarized depth image comprises:
acquiring a hole point and a non-hole point in the original depth image;
and performing binarization processing on the original depth image according to the mode that the hole point is 0 and the non-hole point is 1 to obtain a binarized depth image.
3. The depth image hole filling method based on discrete wavelet decomposition according to claim 1, wherein repairing the first high frequency information image by using the second high frequency information image to obtain a repaired first high frequency information image comprises:
comparing the first high-frequency information image with the second high-frequency information image, and setting zero to the pixel value of the corresponding position in the first high-frequency information image when the pixel value of the same position is not zero so as to obtain a repaired first high-frequency information image.
4. The depth image hole filling method based on discrete wavelet decomposition according to claim 1, wherein the fusing processing of the initial image after hole filling with the original depth image comprises:
traversing the initial image filled with the holes, and carrying out depth filling again on each pixel point, if the corresponding point of the current point in the original depth image is a non-hole point, replacing the depth value of the current point by the depth value of the point in the original depth image; otherwise, the depth value of the current point is reserved.
5. Depth image cavity filling device based on discrete wavelet decomposition, characterized by comprising:
the first transformation module (1) is used for acquiring an original depth image and performing discrete wavelet transformation to obtain a first low-frequency information image and a first high-frequency information image;
the low-frequency filling module (2) is used for filling the cavity points in the first low-frequency information image to obtain a first low-frequency information image with the filled cavities;
the binarization module (3) is used for carrying out binarization processing on the original depth image to obtain a binarized depth image;
a second transformation module (4) for performing discrete wavelet transformation on the binarized depth image to obtain a second low-frequency information image and a second high-frequency information image;
the high-frequency restoration module (5) is used for restoring the first high-frequency information image by utilizing the second high-frequency information image to obtain a restored first high-frequency information image;
the inverse transformation module (6) is used for performing discrete wavelet inverse transformation on the first low-frequency information image after cavity filling and the first high-frequency information image after repairing to obtain an initial image after cavity filling;
the fusion processing module (7) is used for carrying out fusion processing on the initial image filled with the cavity and the original depth image to obtain a final depth image;
wherein the low frequency filling module (2) is specifically configured to:
obtaining a hole point and a non-hole point in the image according to the depth value of each pixel point in the first low-frequency information image;
calculating the similarity weight of the current hole point and each non-hole point in the neighborhood of the current hole point;
calculating a depth estimation value of the current cavity point according to the similarity weight, and filling the current cavity point with the depth estimation value;
traversing the first low-frequency information image to fill all the cavity points, and obtaining the first low-frequency information image after cavity filling.
6. The depth image cavity filling electronic equipment based on discrete wavelet decomposition is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface, the memory complete communication with each other through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any one of claims 1-4 when executing a program stored on a memory.
7. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored therein a computer program which, when executed by a processor, implements the method steps of any of claims 1-4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111329375.9A CN114299076B (en) | 2021-11-10 | 2021-11-10 | Depth image cavity filling method and device based on discrete wavelet decomposition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111329375.9A CN114299076B (en) | 2021-11-10 | 2021-11-10 | Depth image cavity filling method and device based on discrete wavelet decomposition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114299076A CN114299076A (en) | 2022-04-08 |
CN114299076B true CN114299076B (en) | 2023-09-19 |
Family
ID=80964059
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111329375.9A Active CN114299076B (en) | 2021-11-10 | 2021-11-10 | Depth image cavity filling method and device based on discrete wavelet decomposition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114299076B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104780355A (en) * | 2015-03-31 | 2015-07-15 | 浙江大学 | Depth-based cavity repairing method in viewpoint synthesis |
WO2017201751A1 (en) * | 2016-05-27 | 2017-11-30 | 北京大学深圳研究生院 | Hole filling method and device for virtual viewpoint video or image, and terminal |
CN108230268A (en) * | 2016-12-21 | 2018-06-29 | 达索系统公司 | Completion is carried out to image |
CN111179195A (en) * | 2019-12-27 | 2020-05-19 | 西北大学 | Depth image hole filling method and device, electronic equipment and storage medium thereof |
CN112991193A (en) * | 2020-11-16 | 2021-06-18 | 武汉科技大学 | Depth image restoration method, device and computer-readable storage medium |
-
2021
- 2021-11-10 CN CN202111329375.9A patent/CN114299076B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104780355A (en) * | 2015-03-31 | 2015-07-15 | 浙江大学 | Depth-based cavity repairing method in viewpoint synthesis |
WO2017201751A1 (en) * | 2016-05-27 | 2017-11-30 | 北京大学深圳研究生院 | Hole filling method and device for virtual viewpoint video or image, and terminal |
CN108230268A (en) * | 2016-12-21 | 2018-06-29 | 达索系统公司 | Completion is carried out to image |
CN111179195A (en) * | 2019-12-27 | 2020-05-19 | 西北大学 | Depth image hole filling method and device, electronic equipment and storage medium thereof |
CN112991193A (en) * | 2020-11-16 | 2021-06-18 | 武汉科技大学 | Depth image restoration method, device and computer-readable storage medium |
Non-Patent Citations (1)
Title |
---|
牟琦 ; 夏蕾 ; 李占利 ; 李洪安 ; .采用曲率扩散和边缘重建的深度图像空洞修复.西安科技大学学报.2020,(第02期),全文. * |
Also Published As
Publication number | Publication date |
---|---|
CN114299076A (en) | 2022-04-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108198145B (en) | Method and device for point cloud data restoration | |
CN108520536B (en) | Disparity map generation method and device and terminal | |
Min et al. | Depth video enhancement based on weighted mode filtering | |
KR101554241B1 (en) | A method for depth map quality enhancement of defective pixel depth data values in a three-dimensional image | |
US8385630B2 (en) | System and method of processing stereo images | |
EP3488424A1 (en) | Systems and methods for improved surface normal estimation | |
CN111179195B (en) | Depth image cavity filling method and device, electronic equipment and storage medium thereof | |
CN113012210A (en) | Method and device for generating depth map, electronic equipment and storage medium | |
Garcia et al. | Unified multi-lateral filter for real-time depth map enhancement | |
CN113971669A (en) | Three-dimensional detection system applied to pipeline damage identification | |
CN111105451B (en) | Driving scene binocular depth estimation method for overcoming occlusion effect | |
CN111739071A (en) | Rapid iterative registration method, medium, terminal and device based on initial value | |
CN114299076B (en) | Depth image cavity filling method and device based on discrete wavelet decomposition | |
Lu et al. | Depth-based view synthesis using pixel-level image inpainting | |
CN111899277A (en) | Moving object detection method and device, storage medium and electronic device | |
Damjanović et al. | Local stereo matching using adaptive local segmentation | |
McKinnon et al. | A semi-local method for iterative depth-map refinement | |
Sheng et al. | Online temporally consistent indoor depth video enhancement via static structure | |
Shabanian et al. | A new hybrid stereo disparity estimation algorithm with guided image filtering-based cost aggregation | |
CN113744416B (en) | Global point cloud filtering method, equipment and storage medium based on mask | |
CN112270693B (en) | Method and device for detecting motion artifact of time-of-flight depth camera | |
Fan et al. | Collaborative three-dimensional completion of color and depth in a specified area with superpixels | |
CN113065566B (en) | Mismatching removal method, system and application | |
Stentoumis et al. | Implementing an adaptive approach for dense stereo-matching | |
CN107610070A (en) | Free stereo matching process based on three shooting collections |
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 |