CN101833758A - Mono-view depth estimating method and system - Google Patents
Mono-view depth estimating method and system Download PDFInfo
- Publication number
- CN101833758A CN101833758A CN200910126078A CN200910126078A CN101833758A CN 101833758 A CN101833758 A CN 101833758A CN 200910126078 A CN200910126078 A CN 200910126078A CN 200910126078 A CN200910126078 A CN 200910126078A CN 101833758 A CN101833758 A CN 101833758A
- Authority
- CN
- China
- Prior art keywords
- degree
- depth
- deep
- mono
- broad
- 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.)
- Pending
Links
Images
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a mono-view depth estimating method and system. The method comprises the following steps of: firstly, dividing two-dimensional images into a plurality of objects; then sensing a DDR (Depth Dimension Region) in the objects, for example ground or floor. The DDR has a horizontal plane. The depth is distributed to the DDR and to the connecting objects in the DDR.
Description
Technical field
The relevant monoscopic of the present invention (mono-view) degree of depth (depth) estimation is particularly about a kind of ground model (ground model) that is used for the mono-view depth estimation.
Background technology
When three-dimensional body by camera or video camera and projection mapping during to the bidimensional image plane because this kind projection is not exclusive many-one conversion, therefore can lose three-dimensional depth information.In other words, can't decide its degree of depth by the imaging point after the projection.In order to obtain a complete reproduction or approximate three-dimensional performance, must recover or produce these three-dimensional depth informations.
In monoscopic (mono-view or monoscopic) depth estimation, can be by its spatial domain or frequency domain and obtain the degree of depth.Alleged " monoscopic " is meant that images of left and right eyes can obtain identical perspective for a scene and see in this instructions.One of tradition mono-view depth estimating method is to capture the degree of depth by the object of which movement degree, therefore is called " degree of depth (depth-from-motion) is obtained in autokinesis " method.Wherein, the object of tool higher motion degree is endowed less (nearer) degree of depth, the opposite is also true.Another kind of traditional mono-view depth estimating method then is to give big (far away) degree of depth to losing burnt zone (for example background), therefore is called " degree of depth (depth-from-focus-cue) is obtained in the self-focusing prompting " method.Also having a kind of traditional mono-view depth estimating method is the point of crossing (or end point) of detecting vanishing line, and gives big (far away) degree of depth, the opposite is also true near the point the end point.
Because it is very limited from the information that spatial domain-frequency domain obtained of monoscopic, and above-mentioned classic method can't solve all problems of real video signal/image, therefore need the depth estimation method that proposes a kind of novelty badly, in order to handle various monoscopic video signal/images.
Summary of the invention
In view of above-mentioned, one of purpose of the present invention is to propose a kind of ground model that is applicable to mono-view depth estimating method or system, and it can handle various scenes so that correct and various depth information to be provided in the occasion that deep and broad degree zone (depth diffusion region) occurs.
The object of the invention to solve the technical problems realizes by the following technical solutions.According to a kind of monoscopic (mono-view) depth estimation method that the present invention proposes, it may further comprise the steps: bidimensional image is divided into a plurality of objects; Detect deep and broad degree zone (DDR) in these a plurality of objects, it has a surface level; Assign the degree of depth and give this deep and broad degree zone (DDR); And assign the degree of depth and give this deep and broad connection object of spending zone (DDR).
The object of the invention to solve the technical problems also can be applied to the following technical measures to achieve further.
Aforesaid mono-view depth estimating method, wherein said deep and broad degree zone (DDR) is ground, ocean or floor.
Aforesaid mono-view depth estimating method, the degree of depth in wherein said deep and broad degree zone (DDR) are assigned dull (monotonic) method of distribution of employing, increase progressively successively from the bottom up.
It is the degree of depth that is connected the junction of object according to this deep and broad degree zone (DDR) with this that aforesaid mono-view depth estimating method, the degree of depth of wherein said connection object are assigned.
Aforesaid mono-view depth estimating method, the degree of depth of wherein said connection object is same as the degree of depth of this junction.
Aforesaid mono-view depth estimating method, it also comprises a step, in order to the three-dimensional body projection mapping to the bidimensional image plane.
Aforesaid mono-view depth estimating method, it also comprises a step, in order to store or to show that this deep and broad degree zone (DDR) is connected the degree of depth of object with this.
The object of the invention to solve the technical problems also realizes by the following technical solutions.According to a kind of monoscopic (mono-view) depth estimation system that the present invention proposes, it comprises: a cutting unit, in order to bidimensional image is divided into a plurality of objects; One deep and broad degree zone (DDR) detecting unit, in order to detect deep and broad degree zone (DDR) in these a plurality of objects, this deep and broad degree zone (DDR) has a surface level; And regional (DDR) dispatch unit of a deep and broad degree, give this deep and broad degree zone (DDR) in order to assign the degree of depth, and the assignment degree of depth is given the connection object in this deep and broad degree zone (DDR).
The object of the invention to solve the technical problems also can be applied to the following technical measures to achieve further.
Aforesaid mono-view depth estimating system, wherein said deep and broad degree zone (DDR) is ground, ocean or floor.
Aforesaid mono-view depth estimating system, the degree of depth in wherein said deep and broad degree zone (DDR) are assigned dull (monotonic) method of distribution of employing, increase progressively successively from the bottom up.
It is the degree of depth that is connected the junction of object according to this deep and broad degree zone (DDR) with this that aforesaid mono-view depth estimating system, the degree of depth of wherein said connection object are assigned.
Aforesaid mono-view depth estimating system, the degree of depth of wherein said connection object is same as the degree of depth of this junction.
Aforesaid mono-view depth estimating system, it also comprises an input media, in order to the three-dimensional body projection mapping to the bidimensional image plane.
Aforesaid mono-view depth estimating system, it also comprises an output unit, in order to store or to show that this deep and broad degree zone (DDR) is connected the degree of depth of object with this.
The present invention compared with prior art has tangible advantage and beneficial effect.As known from the above, for achieving the above object,, at first, bidimensional image is divided into a plurality of objects according to the embodiment of the invention.Then, in these objects, detect deep and broad degree zone (DDR), for example ground or floor.Deep and broad degree zone (DDR) has a surface level.Assign the degree of depth and give deep and broad degree zone (DDR), for example,, increase progressively successively from the bottom up with dull (monotonic) method of distribution.Assign the degree of depth and give the connection object in deep and broad degree zone (DDR), it is to assign with the degree of depth of the junction that is connected object according to deep and broad degree zone (DDR).For example, task connector spare with deep and broad degree of depth branch of spending the junction in zone (DDR).
By technique scheme, mono-view depth estimating method of the present invention and system have following advantage and beneficial effect at least: the present invention proposes to be applicable to the ground model of monoscopic (mono-view) depth estimation method and system, and it can handle various scenes so that correct and various depth information to be provided in the occasion that deep and broad degree zone (DDR) occurs.
Above-mentioned explanation only is the general introduction of technical solution of the present invention, for can clearer understanding technological means of the present invention, and can be implemented according to the content of instructions, and for above-mentioned and other purposes, feature and advantage of the present invention can be become apparent, below especially exemplified by preferred embodiment, and conjunction with figs., be described in detail as follows.
Description of drawings
Fig. 1 is based on the flow chart of steps of monoscopic (mono-view) the depth estimation method of ground model (ground model) in the embodiment of the invention.
Fig. 2 is the calcspar of the mono-view depth estimating system of the embodiment of the invention.
Fig. 3 is the synoptic diagram in the deep and broad degree zone (DDR) of an image.
100: mono-view depth estimating method 11-18: the step of mono-view depth estimating method
200: mono-view depth estimating system 20: input media
22: cutting unit 24: deep and broad degree zone (DDR) detecting unit
26: the dispatch unit of deep and broad degree zone and connection object
28: output unit 30: the person of playing golf
32: ground (deep and broad degree zone)
Embodiment
Reach technological means and the effect that predetermined goal of the invention is taked for further setting forth the present invention, below in conjunction with accompanying drawing and preferred embodiment, mono-view depth estimating method and its embodiment of system, step, structure, feature and effect thereof that foundation the present invention is proposed are elaborated.
Seeing also shown in Fig. 1,2, is respectively based on the calcspar of the mono-view depth estimating system 200 of the flow chart of steps of monoscopic (mono-view) the depth estimation method 100 of ground model (ground model) and the embodiment of the invention in the embodiment of the invention.
In step 11, input media 20 provides or receives one or many two dimensions (plane) inputs image, in order to as the usefulness of carrying out the image/video signal processing of present embodiment.Input media 20 can be a kind of electrooptical device, in order to the three-dimensional body projection mapping to the bidimensional image plane.In the present embodiment, input media 20 can be a camera, obtains bidimensional image in order to capture; Perhaps can be video camera, in order to obtain several images.In another embodiment, input media 20 can be a pre-procesor, and in order to carry out one or more image processing work, for example image enhancement, image restoration, image analysing computer, image compression or image are synthetic.Moreover input media 20 can more comprise a storage device (for example semiconductor storage or hard disk), in order to store through the handled image of pre-procesor.As previously mentioned, when three-dimensional body projection mapping during to the bidimensional image plane, can lose three-dimensional depth information, therefore, other squares that the mono-view depth estimating system 200 of the embodiment of the invention below will be described in detail in detail are how to be used for handling the bidimensional image that input media 20 is provided.
Next, in step 12, will import image/video signal and be divided into a plurality of zones, object (objects) or fragment to cut apart (segmentation) unit 22.In the present embodiment, " unit " speech is in order to censure a circuit, one section program or its combination.The purpose of cutting apart is in order to change the performance kenel of image, to carry out the assignment of the degree of depth in order to subsequent step.The pixel that is positioned at the same area has similar feature, for example color, intensity (intensity) or texture (texture), and the pixel that is positioned at adjacent area then takes on a different character.The cutting techniques that the execution of step 12 can be adopted tradition or develop in the future.
In step 13, (depth diffusion region, DDR) detecting unit 24 detects deep and broad degree zone (DDR) with deep and broad degree zone.According to the ground model that present embodiment proposed, deep and broad degree zone (DDR) can refer to ground, ocean, floor or other surface levels (horizontal plane).When a surface level has identically when cutting apart feature and having suitable area, promptly may be detected as deep and broad degree zone (DDR).In the illustrated image of Fig. 3, the person of playing golf 30 stands on ground (or meadow) 32.In illustrative image, after the cutting apart of previous step 12, can obtain 30 two objects of the ground 32 and the person of playing golf, ground 32 wherein is a deep and broad degree zone.
(step 14) is promptly assigned the degree of depth with the regional dispatch unit 26 of deep and broad degree and is given deep and broad degree zone (DDR) (step 15) after deep and broad degree zone (DDR) is identified.The degree of depth in deep and broad degree zone (for example ground in the example 32) is assigned can adopt dullness (monotonic) method of distribution, increases progressively successively from the bottom up, and it can be expressed as follows:
DepthDDR (y) ↑ when y ↓
Perhaps
DepthDDR＝k/y
Wherein, k is a constant.
In other embodiments, the regional degree of depth assignment of deep and broad degree can be adopted non-dullness (non-monotonic) method of distribution, for example DepthDDR=k/ (y from the bottom up
2).
Moreover the degree of depth of the object that is connected with deep and broad degree zone (DDR) is assigned, and then can be assigned according to the deep and broad degree regional depth of junction by the regional dispatch unit 26 of deep and broad degree.With Fig. 3 image is example, and the person of playing golf 30 is connected in (or standing on) deep and broad degree zone with pin.Therefore, with deep and broad degree zone (DDR) be positioned at the junction (that is, y
0bj) degree of depth branch task the person of playing golf 30.The assignment of the degree of depth can be expressed as follows:
Depth0bj＝DepthDDR(y
0bj)
In general, when an object stopped (or standing) in deep and broad degree zone (DDR) by a tie point, the degree of depth that then deep and broad degree zone (DDR) is positioned at this connection (combination) point minute was tasked this whole object.
If in step 14, do not identify deep and broad degree zone (DDR), perhaps there is no object and be connected with deep and broad degree zone (DDR), then whole (or part) image can carry out the degree of depth according to tradition (or developing it in the future) degree of depth allocating method and assign.In the flow process of Fig. 1,, detect its prospect and background with step 16, and then assign the degree of depth respectively to prospect and background (step 17) for non-deep and broad degree zone (non-DDR) image.In general, the degree of depth of prospect is less than the degree of depth of background.At last, the depth information that step 15 obtained can be merged (step 18) separately or with depth information that step 17 obtains, thereby obtains ultimate depth map (depth map).
Propose to be applicable to the ground model of monoscopic (mono-view) depth estimation method and system according to the foregoing description, it can handle various scenes so that correct and various depth information to be provided in the occasion that deep and broad degree zone (DDR) occurs.
The above, it only is preferred embodiment of the present invention, be not that the present invention is done any pro forma restriction, though the present invention discloses as above with preferred embodiment, yet be not in order to limit the present invention, any those skilled in the art, in not breaking away from the technical solution of the present invention scope, when the technology contents that can utilize above-mentioned announcement is made a little change or is modified to the equivalent embodiment of equivalent variations, in every case be not break away from the technical solution of the present invention content, according to technical spirit of the present invention to any simple modification that above embodiment did, equivalent variations and modification all still belong in the scope of technical solution of the present invention.
Claims (14)
1. mono-view depth estimating method is characterized in that it may further comprise the steps:
Bidimensional image is divided into a plurality of objects;
Detect deep and broad degree zone in these a plurality of objects, it has a surface level;
Assign the degree of depth and give this deep and broad degree zone; And
Assign the degree of depth and give the connection object in this deep and broad degree zone.
2. mono-view depth estimating method according to claim 1 is characterized in that wherein said deep and broad degree zone is ground, ocean or floor.
3. mono-view depth estimating method according to claim 1 is characterized in that the degree of depth in wherein said deep and broad degree zone is assigned the dull method of distribution of employing, increases progressively successively from the bottom up.
4. mono-view depth estimating method according to claim 1, the degree of depth assignment that it is characterized in that wherein said connection object are the degree of depth that is connected the junction of object according to this deep and broad degree zone with this.
5. mono-view depth estimating method according to claim 4 is characterized in that the degree of depth of wherein said connection object is same as the degree of depth of this junction.
6. mono-view depth estimating method according to claim 1 is characterized in that it also comprises a step, in order to the three-dimensional body projection mapping to the bidimensional image plane.
7. mono-view depth estimating method according to claim 1 is characterized in that it also comprises a step, in order to store or to show that this deep and broad degree zone is connected the degree of depth of object with this.
8. mono-view depth estimating system is characterized in that it comprises:
One cutting unit is in order to be divided into bidimensional image a plurality of objects;
One deep and broad degree area detecting unit, in order to detect deep and broad degree zone in these a plurality of objects, this deep and broad degree zone has a surface level; And
The regional dispatch unit of one deep and broad degree is given this deep and broad degree zone in order to assign the degree of depth, and is assigned the connection object that the degree of depth is given this deep and broad degree zone.
9. mono-view depth estimating system according to claim 8 is characterized in that wherein said deep and broad degree zone is ground, ocean or floor.
10. mono-view depth estimating system according to claim 8 is characterized in that the degree of depth in wherein said deep and broad degree zone is assigned the dull method of distribution of employing, increases progressively successively from the bottom up.
11. mono-view depth estimating system according to claim 8, the degree of depth assignment that it is characterized in that wherein said connection object are the degree of depth that is connected the junction of object according to this deep and broad degree zone with this.
12. mono-view depth estimating system according to claim 11 is characterized in that the degree of depth of wherein said connection object is same as the degree of depth of this junction.
13. mono-view depth estimating system according to claim 8 is characterized in that it also comprises an input media, in order to the three-dimensional body projection mapping to the bidimensional image plane.
14. mono-view depth estimating system according to claim 8 is characterized in that it also comprises an output unit, in order to store or to show that this deep and broad degree zone is connected the degree of depth of object with this.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910126078A CN101833758A (en) | 2009-03-09 | 2009-03-09 | Mono-view depth estimating method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910126078A CN101833758A (en) | 2009-03-09 | 2009-03-09 | Mono-view depth estimating method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101833758A true CN101833758A (en) | 2010-09-15 |
Family
ID=42717819
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910126078A Pending CN101833758A (en) | 2009-03-09 | 2009-03-09 | Mono-view depth estimating method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101833758A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI792381B (en) * | 2021-03-25 | 2023-02-11 | 鈺立微電子股份有限公司 | Image capture device and depth information calculation method thereof |
-
2009
- 2009-03-09 CN CN200910126078A patent/CN101833758A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI792381B (en) * | 2021-03-25 | 2023-02-11 | 鈺立微電子股份有限公司 | Image capture device and depth information calculation method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9811733B2 (en) | Method, apparatus and system for selecting a frame | |
| US8611641B2 (en) | Method and apparatus for detecting disparity | |
| US9811921B2 (en) | Apparatus and method for processing a depth image | |
| US8565512B2 (en) | Method, medium, and system generating depth map of video image | |
| CN107392958B (en) | Method and device for determining object volume based on binocular stereo camera | |
| WO2016176840A1 (en) | Depth/disparity map post-processing method and device | |
| US20100079453A1 (en) | 3D Depth Generation by Vanishing Line Detection | |
| US20100220893A1 (en) | Method and System of Mono-View Depth Estimation | |
| CN103384343A (en) | Image cavity filling method and device thereof | |
| Assuja et al. | 3D coordinate extraction from single 2D indoor image | |
| Yan et al. | Depth map generation for 2d-to-3d conversion by limited user inputs and depth propagation | |
| TW201434010A (en) | Image processor with multi-channel interface between preprocessing layer and one or more higher layers | |
| JP5422538B2 (en) | Image processing apparatus, display apparatus, method and program thereof | |
| TW201907364A (en) | Apparatus and method for processing depth map | |
| CN101833758A (en) | Mono-view depth estimating method and system | |
| Yang et al. | Depth map generation from a single image using local depth hypothesis | |
| Jin et al. | Depth-map driven planar surfaces detection | |
| TWI393070B (en) | Human face model construction method | |
| TW201428680A (en) | Image processing apparatus and foreground extraction method for stereo videos | |
| Wang et al. | Sparse-to-dense depth estimation in videos via high-dimensional tensor voting | |
| Yang et al. | RIFO: Restoring images with fence occlusions | |
| CN104125446A (en) | Depth image optimization processing method and device in the 2D-to-3D conversion of video image | |
| Su et al. | A real-time full-HD 2D-to-3D conversion system using multicore technology | |
| Rzeszutek et al. | Efficient automatic depth estimation for video | |
| TWI825892B (en) | 3d format image detection method and electronic apparatus using the same method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20100915 |