CN104677330A - Small binocular stereoscopic vision ranging system - Google Patents
Small binocular stereoscopic vision ranging system Download PDFInfo
- Publication number
- CN104677330A CN104677330A CN201310625230.2A CN201310625230A CN104677330A CN 104677330 A CN104677330 A CN 104677330A CN 201310625230 A CN201310625230 A CN 201310625230A CN 104677330 A CN104677330 A CN 104677330A
- Authority
- CN
- China
- Prior art keywords
- image
- images
- module
- noise
- binocular
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C3/00—Measuring distances in line of sight; Optical rangefinders
- G01C3/10—Measuring distances in line of sight; Optical rangefinders using a parallactic triangle with variable angles and a base of fixed length in the observation station, e.g. in the instrument
- G01C3/18—Measuring distances in line of sight; Optical rangefinders using a parallactic triangle with variable angles and a base of fixed length in the observation station, e.g. in the instrument with one observation point at each end of the base
Abstract
The invention provides a small binocular stereoscopic vision ranging system. The system comprises calibrating cameras (1), an image acquiring module (2), an image preprocessing module (3), a feature extracting module (4), an image stereoscopic matching module (5) and a three-dimensional rebuilding ranging module (6). The system is characterized in that the modules are sequentially connected; the two cameras are used for synchronously acquiring two images; the images are preprocessed to inhibit noise and highlight usable information; the preprocessing of the image includes the following steps: 1, performing grey processing on the images to reduce information contents of the images; 2, performing smoothing filtering on the images through a gaussian filter to remove the influence of noise on the experiment; 3, detecting the image contours through a canny edge detection operator so as to guarantee simple, convenient and clear angle point detection; 4, guaranteeing clear image features and removing small textures and noise by an algorithm of corrosion and expansion. According to the system, the images are preprocessed to verify the feasibility of binocular ranging.
Description
Technical field
The present invention relates to a kind of computer vision system, particularly a kind of binocular stereo vision range measurement system.
Background technology
Vision is an ancient research topic, the mankind mainly through comprise vision, the sense of hearing, sense of smell and sense of touch four large sensory systems come the external world of perception.Wherein the most complicatedly also the most important thing is vision system, the information that people obtain from the external world more than 80% from vision system.Therefore vision is the important channel that people are familiar with the world, and along with the development of computer technology, people more and more expectation computing machine simulate biological vision system, to realizing more practical function.The three-dimensional information of object in the two-dimensional digital image information regeneration real world obtained by Computer Analysis.
Binocular stereo vision is the important branch of of computer vision, and binocular stereo vision ranging technology is an important application of binocular stereo vision.Ranging technology is divided into active depth measuring and passive ranging two kinds.Active depth measuring must transmit to target object, and this signal is easy to be detected, so just do not accomplish find range disguise, and active depth measuring due to the impact being subject to external environment larger, therefore precision is not often very desirable.Binocular stereo vision range finding is the passive ranging system set up according to principle of parallax, this system is the principle design of human use's binocular perceived distance, namely two when removing from slightly different two angles the same scenery observing objective three-dimensional world, due to the projection of geometrical optics, the picture that the point that observer observes becomes on the eyes retina of left and right two is not on same position, and there is certain deviation, this deviation is exactly binocular parallax, and the actual range of objective object is embodied by parallax.Two video cameras obtain two width images from diverse location to same object shooting, find out the corresponding point in two width images, and through calculating parallax, then recover object in real world middle distance information based on principle of triangulation by matching algorithm.Binocular stereo vision range finding only needs just can obtain quite accurate range information by catching image in the process measured, and need not transmit, this has just had again disguise concurrently when ensureing precision, therefore play increasing effect in each field, and advantage is more and more obvious.
Summary of the invention
The object of this invention is to provide a kind of binocular stereo vision application system based on dual camera, demonstrate the feasibility of binocular range finding.
The object of the present invention is achieved like this:
1, a kind of small capacity double item stereo vision range measurement system, its composition comprises: calibrating camera 1, image capture module 2, image pre-processing module 3, characteristic extracting module 4, image stereo matching module 5, three-dimensional reconstruction range finder module 6, it is characterized in that connecting successively between each comprising modules.
2, a kind of small capacity double item stereo vision range measurement system according to claim 1, is characterized in that improving Harris Corner Detection Algorithm directly calculates two eigenwerts, then directly classifies to it.
3, a kind of small capacity double item stereo vision range measurement system according to claim 1, is characterized in that the displacement between calibrating camera is very little, and the gray-value variation of certain some imaging point in two images in space is very little.
First two width images are synchronously obtained by two cameras.Next need to carry out pre-service with restraint speckle also outstanding useful information to image.The pre-service of image comprises following four steps: the first, and first experiment carries out gray proces to reduce the quantity of information of image to image; The second, utilize Gaussian filter to the smoothing filtering of image to remove the impact of noise on experiment; 3rd, utilize the profile of canny edge detection operator to image to detect, make Corner Detection easier and clear; 4th, utilize corrosion expansion algorithm make the feature of image more clear and remove tiny texture and noise.Through the pre-service of image, Corner Detection and characteristic matching are carried out to gained image.In literary composition, Harris corner detection operator is described in detail, and utilize Harris operator extraction to go out the angle point of profile in two width images, then realize the coupling of the corresponding angle point of two width images based on limit restraint principle, and then obtain parallax, thus recover actual depth information.Whole experiment realizes at vs2010 platform invoke OpenCv2.4.3 built-in function.
The present invention also comprises:
The Harris Corner Detection Algorithm improved directly calculates two eigenwerts, then directly classifies to it, thus avoids calculating Harris response function, also just need not consider the value of α.On the other hand, the Harris Corner Detection Algorithm improved need not carry out non-maxima suppression, but chooses tolerance distance, only has a unique point in tolerance distance, in tolerance distance, only have the most outstanding angle point to be detected, and can control angle point number in image.
First this algorithm chooses the point that has minimax eigenwert, then finds remaining angle point according to minimax eigenwert order successively, and the new angle point of last angle point in tolerance distance is left in the basket.
Because the displacement between two video cameras is very little in the present invention, the gray-value variation of certain some imaging point in two images in space is not very large, for all the more so distant object, so above-mentioned hypothesis can be set up in this experiment.Carry out Gray-scale Matching according to such hypothesis, the calculating related to will be simplified to a great extent, and this also just reduces the complexity of problem to a great extent.
If will to a point in two width images to P
1=[x
1y
11]
tand P
2=[x
2y
21]
tcarry out characteristic matching, the coupling between the picture point in the n × n window centered by these two points can be transformed to.Discrimination principle is: the angle of two vector v 1 and v2 is less, then the matching degree of two points is higher, otherwise, lower.Angle between two vectors can calculate with following formula:
Wherein, θ is the angle of two vectors.When cos θ=1, two vectors have optimum matching.When cos θ=0, it is the poorest coupling.During coupling, certain threshold value should being set, when being greater than a certain threshold value, then thinking match point.
Accompanying drawing explanation
Fig. 1 is binocular stereo vision range finding process flow diagram;
Fig. 2 is the Epipolar geometry of parallel double item stereo vision range measurement system;
Fig. 3 is binocular imaging schematic diagram;
Parallax in the parallel binocular imaging of Fig. 4.
Embodiment
Below in conjunction with accompanying drawing citing, the present invention is described in more detail:
A kind of small capacity double item stereo vision range measurement system, its composition comprises: calibrating camera 1, image capture module 2, image pre-processing module 3, characteristic extracting module 4, image stereo matching module 5, three-dimensional reconstruction range finder module 6, it is characterized in that connecting successively between each comprising modules.
It is characterized in that improving Harris Corner Detection Algorithm directly calculates two eigenwerts, then directly classifies to it.
It is characterized in that the displacement between calibrating camera is very little, the gray-value variation of certain some imaging point in two images in space is very little.
The object of calibrating camera is to set up camera imaging model, determining the inside and outside parameter of video camera, thus determines the corresponding relation of target object between world coordinate system and plane of delineation coordinate system.
Binocular vision system realizes its function, and top priority is exactly will by the two width synchronous images of image capture device acquisition for the physical message of object in this system-computed three-dimensional world.Video camera, video frequency collection card etc. are all the image capture devices relatively commonly used.In order to the image acquisition effect obtained, to reach the object that system accurately carries out testing, need to take into full account the factor such as image definition, resolution, in addition, also all to think better of factors such as the power of nature illumination, the performances of equipment, only take into full account the factor likely existed, the digital picture that can be used for processing of better effects if could have been obtained, for the preprocessing process of ensuing image is laid a solid foundation.
In order to obtain the characteristics of image that can be used for carrying out characteristic matching, need extraction image being carried out to unique point.Before this, need to carry out pre-service and feature point detection and extraction to image, obtain the coordinate of the unique point of image, so that carry out Stereo matching to image.The pretreated idiographic flow of image is as follows:
Determine the same object corresponding point relation on two width images in scene, obtaining parallax with this, is the key of the three-dimensional information determining object in three-dimensional world.In order to realize this target, feature extraction can be carried out to image, then characteristic matching being carried out to it, to obtain parallax.The pixel of image or the pixel set of abstract expression are all said features, and conventional matching process has point-like coupling, wire coupling and Region Matching.In general, the feature of large scale is usually containing abundant information, and relative, number own also just becomes little, and therefore matching speed is very fast, and this is its advantage.But meanwhile, seem very complicated to the extraction of these features also with description, the precision of location is also very undesirable.But, small scale features itself but has higher positioning precision, express description also more simple, although the number of this category feature is a lot, but quantity of information is little, in order to reach higher matching precision, stronger constraint criterion to be found when carrying out characteristic matching, and need corresponding matching strategy.
Stereo matching is carried out to image and between two groups of images, sets up a kind of corresponding relation exactly, epipolar-line constraint relation between two video cameras needs to be asked for draw required fundamental matrix further, there is fundamental matrix, just can obtain more matching double points based on this, thus calculate best epipolar-line constraint relation.Stereo matching can be subject to the impact of several factors, because often there is various noise at occurring in nature, and the factor such as the power of illumination, the visual angle of video camera also all can on mating the certain impact of noise, therefore correct matching algorithm is selected to be the key that the match is successful, otherwise the situation of error hiding may be there is, even occur without coupling phenomenon.
Complete camera calibration, calculated camera intrinsic parameter matrix, and achieve the coupling of stereoscopic image, just can calculate the distance of object relative camera, thus obtain the three-dimensional information of object in space.The precision of system is subject to the impact of several factors, the precision comprising camera calibration and the error brought thus, the length of baseline, and the precision etc. of coupling.
Composition graphs 2, Fig. 3, Fig. 4 illustrate binocular range measurement principle.Fig. 2 is the Epipolar geometry of parallel double item stereo vision range measurement system.In Fig. 2, B is base length, and it is the distance between two video camera photocentres, and namely the camera coordinates of two video cameras ties up to the translational movement in x direction.The such putting position of two video cameras makes the same point of object in three dimensions only have u direction to there is side-play amount under two image coordinate systems, and the difference between u1 and u2 is exactly the expression of parallax under image pixel coordinates system, and v1 and v2 is equal.
If meet aforementioned condition, two polar curves that so two video cameras are corresponding should be conllinear, the region finding match point so just can be made to be compressed on the straight line of one dimension by two dimensional image plane, but under the environment of reality, there is the reasons such as certain distortion in the deviation of putting due to two video cameras or camera lens, these two lines are just similar to the relation of conllinear.
As shown in Figure 3, B is baseline, and in binocular stereo vision range measurement system, the coordinate in world coordinate system of three dimensions point W can by picture planimetric coordinates point x1, x2 and x2, and y2 determines.
In Fig. 4, XZ plane is the plane at two camera lens line places.Using the camera coordinate system of left video camera as world coordinate system, then under world coordinate system, the X-axis coordinate of W point can be expressed as:
Wherein, X
1, Z
1for the coordinate figure of spatial point W X-axis and Z axis under world coordinate system.In like manner, if using the camera coordinate system of right video camera as world coordinate system, then under world coordinate system, the X-axis coordinate of W point can be expressed as:
Because B is base length, and the Z axis coordinate of three dimensions point W is identical under the camera coordinate system of two video cameras, therefore there is following relationship:
X
2=X
1+B
Z
2=Z
1=Z 3
Bring formula 3 into 1 and 2 respectively:
Formula 5 is deducted 4 to obtain:
In formula 6, Z is spatial point and the distance as plane, and such depth information just links up with parallax D, and here, parallax D is x
1with x
2difference, be the difference of the x-axis coordinate under the physical coordinates system of image.The degree of depth depends on looks extent, if can obtain the size of parallax, can calculate the three-dimensional information of object based on this, and the distance of the some distance camera plane in three dimensions also just can determine.
Claims (3)
1. a small capacity double item stereo vision range measurement system, its composition comprises: calibrating camera (1), image capture module (2), image pre-processing module (3), characteristic extracting module (4), image stereo matching module (5), three-dimensional reconstruction range finder module (6), it is characterized in that connecting successively between each comprising modules.
2. a kind of small capacity double item stereo vision range measurement system according to claim 1, is characterized in that improving Harris Corner Detection Algorithm directly calculates two eigenwerts, then directly classifies to it.
3. a kind of small capacity double item stereo vision range measurement system according to claim 1, is characterized in that the displacement between calibrating camera is very little, and the gray-value variation of certain some imaging point in two images in space is very little.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310625230.2A CN104677330A (en) | 2013-11-29 | 2013-11-29 | Small binocular stereoscopic vision ranging system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310625230.2A CN104677330A (en) | 2013-11-29 | 2013-11-29 | Small binocular stereoscopic vision ranging system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104677330A true CN104677330A (en) | 2015-06-03 |
Family
ID=53312700
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310625230.2A Pending CN104677330A (en) | 2013-11-29 | 2013-11-29 | Small binocular stereoscopic vision ranging system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104677330A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005999A (en) * | 2015-08-12 | 2015-10-28 | 北京航空航天大学 | Obstacle detection method for blind guiding instrument based on computer stereo vision |
| CN105403287A (en) * | 2015-10-28 | 2016-03-16 | 南开大学 | High-temperature liquid level measuring device and measuring method based on double-camera |
| CN105651258A (en) * | 2015-12-30 | 2016-06-08 | 杨正林 | Initiative-view-angle binocular vision ranging system and initiative-view-angle binocular vision ranging method |
| CN106052637A (en) * | 2016-06-03 | 2016-10-26 | 用友网络科技股份有限公司 | Distance induction method based on dual cameras |
| CN106767716A (en) * | 2016-12-13 | 2017-05-31 | 云南电网有限责任公司电力科学研究院 | High-tension bus-bar range-measurement system and method based on FPGA hardware and binocular vision |
| CN107289869A (en) * | 2017-06-08 | 2017-10-24 | 杭州联络互动信息科技股份有限公司 | A kind of method, apparatus and system that 3D measurements are carried out using matrix camera lens |
| CN107560592A (en) * | 2017-08-21 | 2018-01-09 | 河南中光学集团有限公司 | A kind of precision ranging method for optronic tracker linkage target |
| CN107688174A (en) * | 2017-08-02 | 2018-02-13 | 北京纵目安驰智能科技有限公司 | A kind of image distance-finding method, system, storage medium and vehicle-mounted visually-perceptible equipment |
| CN108982554A (en) * | 2018-06-20 | 2018-12-11 | 国网河南省电力公司电力科学研究院 | A kind of unmanned plane lifting x-ray detector and detection method |
| CN109919247A (en) * | 2019-03-18 | 2019-06-21 | 北京石油化工学院 | Characteristic point matching method, system and equipment in harmful influence stacking binocular ranging |
| CN110110131A (en) * | 2019-05-23 | 2019-08-09 | 北京航空航天大学 | It is a kind of based on the aircraft cable support of deep learning and binocular stereo vision identification and parameter acquiring method |
| CN111473767A (en) * | 2020-04-16 | 2020-07-31 | 福建汇川物联网技术科技股份有限公司 | Remote distance measurement method and device |
| CN112014393A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method based on target visual effect |
| CN112016558A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method based on image quality |
| CN112017243A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method |
| WO2021022934A1 (en) * | 2019-08-05 | 2021-02-11 | 上海亨临光电科技有限公司 | Passive millimeter wave/terahertz imaging technology-based three-dimensional imaging method |
| CN113081270A (en) * | 2021-03-09 | 2021-07-09 | 北京铸正机器人有限公司 | Tail end clamping device for robotic surgery |
-
2013
- 2013-11-29 CN CN201310625230.2A patent/CN104677330A/en active Pending
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005999A (en) * | 2015-08-12 | 2015-10-28 | 北京航空航天大学 | Obstacle detection method for blind guiding instrument based on computer stereo vision |
| CN105005999B (en) * | 2015-08-12 | 2018-08-10 | 北京航空航天大学 | It is a kind of based on obstacle detection method of the computer stereo vision towards apparatus for guiding blind |
| CN105403287A (en) * | 2015-10-28 | 2016-03-16 | 南开大学 | High-temperature liquid level measuring device and measuring method based on double-camera |
| CN105651258A (en) * | 2015-12-30 | 2016-06-08 | 杨正林 | Initiative-view-angle binocular vision ranging system and initiative-view-angle binocular vision ranging method |
| CN105651258B (en) * | 2015-12-30 | 2018-07-13 | 杨正林 | Active visual angle binocular range-measurement system and method |
| CN106052637B (en) * | 2016-06-03 | 2018-12-18 | 用友网络科技股份有限公司 | Based on dual camera apart from inducing method |
| CN106052637A (en) * | 2016-06-03 | 2016-10-26 | 用友网络科技股份有限公司 | Distance induction method based on dual cameras |
| CN106767716A (en) * | 2016-12-13 | 2017-05-31 | 云南电网有限责任公司电力科学研究院 | High-tension bus-bar range-measurement system and method based on FPGA hardware and binocular vision |
| CN107289869A (en) * | 2017-06-08 | 2017-10-24 | 杭州联络互动信息科技股份有限公司 | A kind of method, apparatus and system that 3D measurements are carried out using matrix camera lens |
| CN107688174A (en) * | 2017-08-02 | 2018-02-13 | 北京纵目安驰智能科技有限公司 | A kind of image distance-finding method, system, storage medium and vehicle-mounted visually-perceptible equipment |
| CN107560592B (en) * | 2017-08-21 | 2020-08-18 | 河南中光学集团有限公司 | Precise distance measurement method for photoelectric tracker linkage target |
| CN107560592A (en) * | 2017-08-21 | 2018-01-09 | 河南中光学集团有限公司 | A kind of precision ranging method for optronic tracker linkage target |
| CN108982554A (en) * | 2018-06-20 | 2018-12-11 | 国网河南省电力公司电力科学研究院 | A kind of unmanned plane lifting x-ray detector and detection method |
| CN109919247A (en) * | 2019-03-18 | 2019-06-21 | 北京石油化工学院 | Characteristic point matching method, system and equipment in harmful influence stacking binocular ranging |
| CN110110131A (en) * | 2019-05-23 | 2019-08-09 | 北京航空航天大学 | It is a kind of based on the aircraft cable support of deep learning and binocular stereo vision identification and parameter acquiring method |
| CN110110131B (en) * | 2019-05-23 | 2021-04-13 | 北京航空航天大学 | Airplane cable support identification and parameter acquisition method based on deep learning and binocular stereo vision |
| WO2021022934A1 (en) * | 2019-08-05 | 2021-02-11 | 上海亨临光电科技有限公司 | Passive millimeter wave/terahertz imaging technology-based three-dimensional imaging method |
| CN111473767A (en) * | 2020-04-16 | 2020-07-31 | 福建汇川物联网技术科技股份有限公司 | Remote distance measurement method and device |
| CN112014393A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method based on target visual effect |
| CN112016558A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method based on image quality |
| CN112017243A (en) * | 2020-08-26 | 2020-12-01 | 大连信维科技有限公司 | Medium visibility identification method |
| CN112014393B (en) * | 2020-08-26 | 2023-12-19 | 大连信维科技有限公司 | Medium visibility recognition method based on target visual effect |
| CN113081270A (en) * | 2021-03-09 | 2021-07-09 | 北京铸正机器人有限公司 | Tail end clamping device for robotic surgery |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104677330A (en) | Small binocular stereoscopic vision ranging system | |
| CN103528571B (en) | Single eye stereo vision relative pose measuring method | |
| CN103115613B (en) | Three-dimensional space positioning method | |
| CN110826499A (en) | Object space parameter detection method and device, electronic equipment and storage medium | |
| CN105043350A (en) | Binocular vision measuring method | |
| CN104574393A (en) | Three-dimensional pavement crack image generation system and method | |
| CN111047634B (en) | Scene depth determination method, device, equipment and storage medium | |
| CN102997891A (en) | Device and method for measuring scene depth | |
| EP3953903A1 (en) | Scale-aware monocular localization and mapping | |
| CN105277144A (en) | Land area rapid detection method based on binocular vision and detection device thereof | |
| CN110096993A (en) | The object detection apparatus and method of binocular stereo vision | |
| CN106500625A (en) | A kind of telecentricity stereo vision measuring apparatus and its method for being applied to the measurement of object dimensional pattern micron accuracies | |
| CN111640156A (en) | Three-dimensional reconstruction method, equipment and storage equipment for outdoor weak texture target | |
| CN107374638A (en) | A kind of height measuring system and method based on binocular vision module | |
| KR20110055032A (en) | Apparatus and method for generating three demension content in electronic device | |
| CN105335934B (en) | Disparity map computational methods and device | |
| CN103544492A (en) | Method and device for identifying targets on basis of geometric features of three-dimensional curved surfaces of depth images | |
| CN110030979A (en) | A kind of space non-cooperative target relative pose measurement method based on sequence image | |
| CN110570468A (en) | Binocular vision depth estimation method and system based on depth learning | |
| CN115731345A (en) | Human body three-dimensional reconstruction method based on binocular vision | |
| CN104296690A (en) | Multi-line structure light three-dimensional measuring method based on image fusion | |
| KR20160039447A (en) | Spatial analysis system using stereo camera. | |
| Iqbal et al. | Real time 3D depth estimation and measurement of un-calibrated stereo and thermal images | |
| CN106303501A (en) | Stereo-picture reconstructing method based on image sparse characteristic matching and device | |
| KR20120056668A (en) | Apparatus and method for recovering 3 dimensional information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150603 |