CN104952099A - One-house-one-scene digital house seeing system - Google Patents
One-house-one-scene digital house seeing system Download PDFInfo
- Publication number
- CN104952099A CN104952099A CN201410120662.2A CN201410120662A CN104952099A CN 104952099 A CN104952099 A CN 104952099A CN 201410120662 A CN201410120662 A CN 201410120662A CN 104952099 A CN104952099 A CN 104952099A
- Authority
- CN
- China
- Prior art keywords
- key frame
- information
- difference degree
- house
- image
- 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
Abstract
The invention discloses a one-house-one-scene digital house seeing system. Prior two-dimensional image information is inputted firstly, a two-dimensional key frame is defined randomly, foreground subtraction is carried out on the key frame, and related basic information of a main target in the key frame is marked during a simple man-machine interaction process; when the key frame image is segmented, difference information of the related target object is generated; after the core information of the key frame is marked, the system can generate a difference graph of the key frame so as to generate a three-dimensional image frame corresponding to the key frame, difference graphs of other non-key frames are automatically generated according to the difference graph of the key frame, and a house type presentation three-dimensional image is outputted finally. Through carrying out informational three-dimensional processing on the two-dimensional image, the scene outside the window and an indoor pattern can be simulated truthfully, a house purchaser can know conditions of the house type in all directions, the purchasing cost of the client is reduced, and the marketing degree is high.
Description
Technical field
The invention belongs to field of information processing, more specifically a kind of digital house viewing system of comprehensive displaying one room one scape of information process-.
Background technology
Today of digital development, all information displays can be represented by computing machine panorama, and the virtual model room that the three-dimensional visualization that present stage grows up selects house system to be integrated with can arbitrarily be viewed and admired, house purchaser passes through the manipulation of keyboard and mouse, the roaming of virtual show house multi-angle can be realized visit, strong body faces on-the-spot sensation, allows house purchaser in interactive operation, experience the warmth of " family " in the future.Landscape house is the important component part of Value of house, and three-dimensional visualization selects house system by carrying out real simulation to view outside window, and house purchaser can stand in balcony or the window limit of virtual show house, appreciates graceful garden landscape true to nature, experiences the value of landscape house in advance! Three-dimensional visualization selects this exclusive function of house system by the location value of building, landscape dominance expressively incisively and vividly.
Three-dimensional house type picture viewing system displays the house type model after fineness finishing with three-dimensional pattern the most true to nature, system comprises: house type detailed dimensions information displaying, sunlight are towards the function such as, bearing wall display, 360 degree of rotary viewings, Scalable, direction displays.Based on these functions, house purchaser can understand house type general layout easily, intuitively, and can reasonably anticipation future finishing view.Interactive three-dimensional sand table display systems, will plan to build building 1:1 and accurately emulate on computer platform, comprise the functions such as panorama is got a bird's eye view, surrounding enviroment displaying.Building information specifically can be understood by shirtsleeve operation, building is represented more intuitively.
In prior art, there are the following problems: three-dimensional display true development, needs high-quality making apparatus, but somewhat expensive, under the prerequisite having high-quality equipment, how about produce the program of the most comfortable viewing, need the renewal of ideas, manufacture skill and making experience.Have expert to point out, the talent and intention are the basic of three-dimensional industry benign development, domestic current in great shortage three-dimensional professional.Three-dimensional television content is deficient, three-dimensional standard disunity again, even if there is advanced terminal device, but coupling becomes again problem etc.
Summary of the invention
1, object of the present invention.
The room one scape house viewing system that the present invention proposes, by after converting two-dimensional image information to 3-D view, be supplied to the experience of the displaying house type of client's omnibearing stereo, the sharpness of 3-D view can be improved, the problem that the existing conversion equipment of solution is very expensive by improvement of the present invention.
2, the technical solution adopted in the present invention.
One room one scape numeral house viewing system, input priori two-dimensional image information information, first two-dimentional key frame is defined at random, key frame is done front background separation, and marked by the relevant essential information of simple interactive process to main target in key frame, when key frame images is divided complete after, the difference degree information of related objective object can generate, after the core information of key frame marks, system can generate the difference degree figure of key frame thus be used for generating three dimensional image frames corresponding to key frame, the difference degree figure of other non-key frame is by the generation of the difference degree figure robotization according to key frame, the acquisition of key frame difference degree information and based on after the non-key difference degree information iteration of key frame difference degree information, system is by according to the difference degree information of key frame and the difference degree figure non-key with the generation of relativeness information dynamic autoization of key frame, finally export as house type shows 3-D view.
Preferred described change three-dimensional based on difference degree figure and adopt with the following method: gray scale 0 represents solstics, and gray scale 255 represents closest approach.The corresponding unique width two dimensional image of each difference degree figure, with difference degree figure be the situation of equal interval quantizing for representative, in image, the difference degree of all pixels can be obtained by following formulae discovery:
Wherein
,
, represent in image farthest with the difference of closest approach,
formulate the gray level of all depth maps.
In a specific embodiment of the present invention, preferably
value be 255.
In another specific embodiment of the present invention, before described priori two-dimensional image information, first pre-service is carried out to two-dimensional image information, the value of pixel any in image is recalculated according to the mean value getting surrounding pixel, determine center pixel value further by the difference introducing center pixel and surrounding pixel, specifically calculate according to following formula:
Wherein
for determining the weights of pixel weight, all pixel weights add and are 1,
computing method as follows:
with
these two parameters represent
repetition filter times,
it is space Gauss's weight.
In another specific embodiment of the present invention, described front background separation, classifies two dimensional image, defines the center of fiqure of each class, mark center of fiqure position; Then each point is mated with nearest center of fiqure, at image be a little all matched to nearest center of fiqure, then calculate the center of gravity of each class, pending image or image are divided into foreground pixel and background pixel clearly, and background pixel value are set as constant or zero.Preferred optimal classification number is 2.
In another specific embodiment of the present invention, correct described prospect difference degree, target image is longitudinal layered, and then object mean difference generates, and finally distributes the difference of target object pixel.
3, beneficial effect of the present invention.
(1) the pretreated method of application image carries out to two dimensional image the error rate that pre-service effectively reduces D reconstruction step.
(2) the present invention uses sorting algorithm to classify to the image after pretreated, thus obtains the image of the imparting difference degree that prospect background is separated.
The present invention is by carrying out information-based 3D processing to two dimensional image, and can carry out real simulation to view and indoor layout outside window, house purchaser understands the situation of house type in all directions, reduces the acquisition cost of client, and marketing degree is higher.
Accompanying drawing explanation
Fig. 1 house type shows processing logic block diagram.
Embodiment
In order to enable the auditor of Patent Office especially the public clearly understand technical spirit of the present invention and beneficial effect, applicant will elaborate below by way of example, but be not all the restriction to the present invention program to the description of embodiment, any conceive according to the present invention done be only pro forma but not substantial equivalent transformation and all should be considered as technical scheme category of the present invention.
Embodiment 1
Two-dimensional image turns 3-D technology totally can be divided into two large classes: 1) man-machine interaction class, this type of D reconstruction technology needs technician when processing two-dimentional film source, the prior imformation necessary to correlation procedure input, thus make computer based in these prior imformation generating three-dimensional figures as sequence.2) non-man-machine interactive class (robotization), this type of technology only need using two-dimensional image sequence as input, without any need for prior imformation or prior imformation Already in internal system, conversion system can full automatic generating three-dimensional figures as sequence.Obviously, from the angle of quantity of information, compare automatic creation system, the D reconstruction technology of man-machine interaction must be that effect is good when transforming.
The room one scape numeral house viewing system that the present invention proposes, key frame is done front background separation, and marked by the relevant essential information of simple interactive process to main target in key frame, when key frame images is divided complete after, the difference degree information of related objective object can generate.Then, we have proposed a kind of image three-dimensional information reconstruction system of based target tracking technique, system is by according to the difference degree information of key frame and the difference degree figure non-key with the generation of relativeness information dynamic autoization of key frame.Finally, necessary three-dimensional information all generates.
Based on transmission, the procedure for displaying of the encoding and decoding of difference degree figure rendering technique 3-D view.
Gray scale 0 represents solstics, and gray scale 255 represents closest approach.The corresponding unique width two dimensional image of each difference degree figure, with preferred described the changing three-dimensional based on difference degree figure and adopt with the following method of difference: gray scale 0 represents solstics, and gray scale 255 represents closest approach.The corresponding unique width two dimensional image of each difference degree figure, with difference degree figure be the situation of equal interval quantizing for representative, in image, the difference degree of all pixels can be obtained by following formulae discovery:
Wherein
,
, represent in image farthest with the difference of closest approach,
formulate the gray level of all depth maps,
value be generally 255.
Embodiment 2
House type as shown in Figure 1 shows processing logic block diagram, before described priori two-dimensional image information, first pre-service is carried out to two-dimensional image information, noise in two dimensional image before treatment for D reconstruction effectively can be removed, to make to generate difference degree information accurately in image after D reconstruction process, reduce error rate, thus level and smooth, denoising effect can be ensured, can well keep image border again.
The value of pixel any in image is recalculated according to the mean value getting surrounding pixel, determines center pixel value further by the difference introducing center pixel and surrounding pixel, specifically calculate according to following formula:
Wherein
for determining the weights of pixel weight, all pixel weights add and are 1,
computing method as follows:
with
these two parameters represent
repetition filter times,
it is space Gauss's weight.
Front background separation, carries out point 2 classes by two dimensional image, defines the center of fiqure of each class, mark center of fiqure position; Then mated with nearest center of fiqure by each point, the position that each center of fiqure is placed should not be excessively near, will spread all over whole image fully.At image be a little all matched to nearest center of fiqure, then calculate the center of gravity of each class, pending image or image are divided into foreground pixel and background pixel clearly, and background pixel value is set as constant or zero.
Although have passed through operation above, the prospect of image has well been separated with background and has been carried out a step difference degree assignment, but for the difference degree figure wanting to realize 3-D display or inadequate, this is because one should be unique or without excessive gap in the difference angle value of the actual object of prospect.So will correct prospect difference degree, for this reason, although prospect and background are all separated, be necessary to carry out unitized process to the difference angle value of foreground target object.And the difference angle value of giving more accurately reflect that the actual relative difference degree position at object place is better.Target image is longitudinal layered, and the difference that then object is average generates, and the difference of ideal pixel is distributed, thus realizes effectively correcting.
Principle of work of the present invention:
Follow the room one scape numeral house viewing system based on difference degree figure renders three-dimensional two-dimensional image image conversion 3-D view.First pre-service is carried out to image, comprise two dimensional image acquisition and the pre-service to image.Use result surface, the pretreated method of application image carries out to two dimensional image the error rate that pre-service effectively reduces D reconstruction step herein; We use sorting algorithm to classify to the image after pretreated, thus obtain the image that difference degree is given in wait that prospect background is separated.In order to give difference degree to the image pixel after clustering algorithm process more accurately, use based on experimental image layout model, for making difference degree assignment to target each in image is distinctive.Result shows, put forward mode and provide the possible ways that a kind of two-dimensional image turns three-dimensional, in systems in which, propose the difference degree figure automatic generating calculation that based target track algorithm completes non-key frame.System can be used for the difference degree information needed for generating three-dimensional figures picture, first by image zooming-out key frame, carries out Iamge Segmentation to key frame.Mutual afterwards the main target of key frame to be marked, the difference degree information assignment thereupon of target-marking.For non-key frame, according to the difference degree figure of key frame, the generation of the non-key frame image difference degree automatic generating calculation robotization that the based target using us to propose is followed the tracks of, the difference degree image sequence corresponding to two-dimensional image sequence all generates, and last generating three-dimensional figures is shown as house type.
Claims (7)
1. a room one scape numeral house viewing system, it is characterized in that: input priori two-dimensional image information information, first two-dimentional key frame is defined at random, key frame is done front background separation, and marked by the relevant essential information of simple interactive process to main target in key frame, when key frame images is divided complete after, the difference degree information of related objective object can generate, after the core information of key frame marks, system can generate the difference degree figure of key frame thus be used for generating three dimensional image frames corresponding to key frame, the difference degree figure of other non-key frame is by the generation of the difference degree figure robotization according to key frame, the acquisition of key frame difference degree information and based on after the non-key difference degree information iteration of key frame difference degree information, system is by according to the difference degree information of key frame and the difference degree figure non-key with the generation of relativeness information dynamic autoization of key frame, finally export as house type shows 3-D view.
2. a room one scape numeral house viewing system according to claim 1, it is characterized in that: described change three-dimensional based on difference degree figure and adopt with the following method: gray scale 0 represents solstics, gray scale 255 represents closest approach, the corresponding unique width two dimensional image of each difference degree figure, with difference degree figure be the situation of equal interval quantizing for representative, in image, the difference degree of all pixels can be obtained by following formulae discovery:
Wherein
,
, represent in image farthest with the difference of closest approach,
formulate the gray level of all depth maps.
3. a room one scape numeral house viewing system according to claim 2, is characterized in that: preferably
value be 255.
4. a room one scape numeral house viewing system according to claim 1, it is characterized in that: before described priori two-dimensional image information, first pre-service is carried out to two-dimensional image information, the value of pixel any in image is recalculated according to the mean value getting surrounding pixel, determine center pixel value further by the difference introducing center pixel and surrounding pixel, specifically calculate according to following formula:
Wherein
for determining the weights of pixel weight, all pixel weights add and are 1,
computing method as follows:
with
these two parameters represent
repetition filter times,
it is space Gauss's weight.
5. a room one scape numeral house viewing system according to claim 1, is characterized in that: described front background separation, is classified by two dimensional image, define the center of fiqure of each class, mark center of fiqure position; Then each point is mated with nearest center of fiqure, at image be a little all matched to nearest center of fiqure, then calculate the center of gravity of each class, pending image or image are divided into foreground pixel and background pixel clearly, and background pixel value are set as constant or zero.
6. a room one scape numeral house viewing system according to claim 1, is characterized in that: optimal classification number is 2.
7. the room one scape numeral house viewing system according to claim 5 or 6, it is characterized in that: correct described prospect difference degree, target image is longitudinal layered, and then object mean difference generates, and finally distributes the difference of target object pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410120662.2A CN104952099A (en) | 2014-03-28 | 2014-03-28 | One-house-one-scene digital house seeing system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410120662.2A CN104952099A (en) | 2014-03-28 | 2014-03-28 | One-house-one-scene digital house seeing system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104952099A true CN104952099A (en) | 2015-09-30 |
Family
ID=54166725
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410120662.2A Pending CN104952099A (en) | 2014-03-28 | 2014-03-28 | One-house-one-scene digital house seeing system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104952099A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488230A (en) * | 2016-01-22 | 2016-04-13 | 国家电网公司 | Synchronous house selecting device capable of supporting remote off-site and multiple-point selection as well as method |
| CN110798610A (en) * | 2018-08-02 | 2020-02-14 | 光宝电子(广州)有限公司 | Image display system and image display method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101930627A (en) * | 2010-09-10 | 2010-12-29 | 西安新视角信息科技有限公司 | Three-dimensional dwelling size modeling method based on two-dimensional dwelling size diagram |
| US20130127989A1 (en) * | 2011-11-22 | 2013-05-23 | Sonic Solutions, Llc | Conversion of 2-Dimensional Image Data into 3-Dimensional Image Data |
-
2014
- 2014-03-28 CN CN201410120662.2A patent/CN104952099A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101930627A (en) * | 2010-09-10 | 2010-12-29 | 西安新视角信息科技有限公司 | Three-dimensional dwelling size modeling method based on two-dimensional dwelling size diagram |
| US20130127989A1 (en) * | 2011-11-22 | 2013-05-23 | Sonic Solutions, Llc | Conversion of 2-Dimensional Image Data into 3-Dimensional Image Data |
Non-Patent Citations (1)
| Title |
|---|
| 蒋昊: "三维显示信息重建及评价方法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488230A (en) * | 2016-01-22 | 2016-04-13 | 国家电网公司 | Synchronous house selecting device capable of supporting remote off-site and multiple-point selection as well as method |
| CN105488230B (en) * | 2016-01-22 | 2017-03-29 | 国家电网公司 | Room device and method is selected in a kind of synchronization for supporting long-range strange land multiple spot |
| CN110798610A (en) * | 2018-08-02 | 2020-02-14 | 光宝电子(广州)有限公司 | Image display system and image display method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104685858B (en) | Immersive videoconference method and system | |
| CN108027984B (en) | Method and system for detecting and combining structural features in 3D reconstruction | |
| Karsch et al. | Depth transfer: Depth extraction from video using non-parametric sampling | |
| US8644596B1 (en) | Conversion of monoscopic visual content using image-depth database | |
| Feng et al. | Object-based 2D-to-3D video conversion for effective stereoscopic content generation in 3D-TV applications | |
| Shao et al. | QoE-guided warping for stereoscopic image retargeting | |
| CN103345771A (en) | Efficient image rendering method based on modeling | |
| CN103929634A (en) | 3d-animation Effect Generation Method And System | |
| CN102609950A (en) | Two-dimensional video depth map generation process | |
| Du et al. | Video fields: fusing multiple surveillance videos into a dynamic virtual environment | |
| CN104217461B (en) | A parallax mapping method based on a depth map to simulate a real-time bump effect | |
| Suenaga et al. | A practical implementation of free viewpoint video system for soccer games | |
| Wang et al. | Comfort-driven disparity adjustment for stereoscopic video | |
| Sandnes | Sketching 3D immersed experiences rapidly by hand through 2D cross sections | |
| CN103514624A (en) | Method for estimating quantity of light received by participating media, and corresponding device | |
| CN111292425A (en) | View synthesis method based on monocular and binocular mixed data set | |
| CN104952099A (en) | One-house-one-scene digital house seeing system | |
| CN108616746A (en) | The method that 2D panoramic pictures based on deep learning turn 3D panoramic pictures | |
| Dai et al. | PBR-Net: Imitating physically based rendering using deep neural network | |
| Yang et al. | Depth map generation using local depth hypothesis for 2D-to-3D conversion | |
| CN104243950B (en) | For the method and apparatus that 2 dimension contents are converted to 3-dimensional content in real time | |
| CN112002019B (en) | Method for simulating character shadow based on MR mixed reality | |
| Wang et al. | An intelligent screen system for context-related scenery viewing in smart home | |
| Chen et al. | A quality controllable multi-view object reconstruction method for 3D imaging systems | |
| Liu et al. | See360: Novel panoramic view interpolation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Four 215513 Jiangsu Sea city of Suzhou province Changshou City economic and Technological Development Zone No. 11 Building No. 1 Room 405 Branch Park Applicant after: SUZHOU MAIN GOAL SOFTWARE TECHNOLOGY CO., LTD. Address before: Four 215513 Jiangsu Sea city of Suzhou province Changshou City economic and Technological Development Zone No. 11 Building No. 1 Room 405 Branch Park Applicant before: Suzhou Mei Gu Shi Dian software Science and Technology Ltd. |
|
| COR | Change of bibliographic data | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150930 |