CN109697745A - Barrier perspective method and barrier arrangement for perspective - Google Patents

Barrier perspective method and barrier arrangement for perspective Download PDF

Info

Publication number
CN109697745A
CN109697745A CN201711001278.0A CN201711001278A CN109697745A CN 109697745 A CN109697745 A CN 109697745A CN 201711001278 A CN201711001278 A CN 201711001278A CN 109697745 A CN109697745 A CN 109697745A
Authority
CN
China
Prior art keywords
image
barrier
width
reserved
perspective
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
Application number
CN201711001278.0A
Other languages
Chinese (zh)
Inventor
王蕙雯
何其勋
李学文
辛怡德
郭俊彦
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Yuzhan Precision Technology Co ltd
Hon Hai Precision Industry Co Ltd
Original Assignee
Shenzhen Yuzhan Precision Technology Co ltd
Hon Hai Precision Industry Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Yuzhan Precision Technology Co ltd, Hon Hai Precision Industry Co Ltd filed Critical Shenzhen Yuzhan Precision Technology Co ltd
Priority to CN201711001278.0A priority Critical patent/CN109697745A/en
Publication of CN109697745A publication Critical patent/CN109697745A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/002D [Two Dimensional] image generation
    • G06T11/60Editing figures and text; Combining figures or text
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/50Context or environment of the image
    • G06V20/56Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G06V20/58Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Studio Devices (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a kind of barrier perspective methods.The barrier perspective method is comprising steps of acquire at least two images that scheduled barrier section blocks from least two different perspectivess;The part blocked in at least two images by barrier is deleted, at least two width is generated and reserves image;Determine that at least two width reserve the stitching position of image;Splice at least two width and reserves image.The present invention further discloses corresponding barrier arrangement for perspective.Barrier perspective method and barrier arrangement for perspective disclosed by the invention are intended to blind area caused by reducing or eliminating barrier.

Description

Barrier perspective method and barrier arrangement for perspective
Technical field
The present invention relates to a kind of barrier perspective method and barrier arrangement for perspective.
Background technique
It in the presence of having barrier, tends to form blind area, observer or filming apparatus can not obtain The visual field of shelter stop portions, so that complete picture can not be obtained.For example, windshield on automobile and front door it Between column play a supportive role and be called A column, A column causes to regard due to that can block driver part sight in the process of moving Feel blind area.The presence of pillar A blind is to cause the one of the major reasons of traffic accident, and sight is blocked inconvenient especially in driver turn That embodies during curved or lane-change becomes apparent, so the narrower A column theoretically the better, and A column supports automobile front weight It is required that there is certain intensity, therefore causes barrier and can not eliminate contradiction between the integrality demand of the visual field.
Notification number is that the Taiwan Patent of TWI267061 discloses a kind of method of processing multi-layer image picture overlapping, the party Whether method comprises the steps of the shielding weight of one first picture data of detecting between a preset range;And when first picture When the shielding weight system of face data is in the preset range, according to first picture data, one second picture data, and it is somebody's turn to do The shielding weight of second picture data generates a third picture data.The disclosure of TWI267061 provide a kind of multilayer picture optical with The method of translucent form overlapping display, however blind area caused by it still fails for reducing or eliminating barrier.
Summary of the invention
In view of this, it is necessary to provide a kind of barrier perspective method and barrier arrangement for perspective, to reduce or eliminate barrier Hinder blind area caused by object.
A kind of barrier perspective method hides comprising steps of acquiring scheduled barrier section from least two different perspectivess At least two images of gear;The part blocked in at least two images by barrier is deleted, the reserved figure of at least two width is generated Picture;Determine that at least two width reserve the stitching position of image;Splice at least two width and reserves image.
Preferably, adjacent two visual angle has overlapping region at least two different perspectives.
Preferably, determining that the step of at least two width reserve the stitching position of image specifically includes: according to it is described extremely The pre-set dimension of few two different perspectivess and at least two images determines the location information of at least two images;It compares The location information of adjacent image is referring to calculating angular transformation amount and translation transformation amount with a wherein image.
Preferably, determining that the step of at least two width reserve the stitching position of image specifically includes: to adjacent pre- Image is stayed to carry out feature extraction;It compares the feature extracted and carries out image registration.
Preferably, described compare the feature extracted and specifically include the step of carrying out image registration: according to adjacent image The position of middle lap determines the transformation relation of two images, finds out the transformation model of mapping relations between image to be spliced; The registration of multiple image is realized in the way of one or more registrations.
Preferably, being further comprised the steps of: reserved with a wherein width before splicing the step of at least two width reserve image Image is reference, reserves other reserved images in image at least two width and carries out size change over.
Preferably, being further comprised the steps of: reserved with a wherein width before splicing the step of at least two width reserve image Image is reference, reserves other reserved images in image at least two width and carries out resolution conversion.
It is specifically included preferably, splicing the step of at least two width reserve image: to pair of adjacent reserved image Sub-region right is answered to merge, which is realized using the blending algorithm of Pixel-level.
A kind of barrier arrangement for perspective, including a photographic device and a processing unit, the photographic device include at least two The direction of group camera, at least two groups camera is different;The signal output end of at least two groups camera is connected to institute State processing unit;After the processing unit receives at least two groups camera shooting information of at least two groups camera, by described at least Two groups of camera shooting information fusions, form the panoramic information of a presumptive area.
Preferably, the visual angle of two adjacent groups camera has overlapping region in at least two groups camera.
It is scheduled by being acquired from least two different perspectivess in above-mentioned barrier perspective method and barrier arrangement for perspective After at least two images that barrier section blocks, the part blocked in at least two images by barrier is deleted, is generated At least two width reserve image, and splice at least two width and reserve image, then at least two width are reserved in image, are not hidden by barrier The part of gear will be summed it up, to obtain the image that a width partially or completely eliminates shield portions, realize perspective barrier Purpose, and then to reduce or eliminate blind area caused by barrier.
Detailed description of the invention
Fig. 1 is structural schematic diagram of the barrier arrangement for perspective in a preferred embodiment.
Fig. 2 is the operation principle schematic diagram of the barrier arrangement for perspective in Fig. 1.
Fig. 3 is operation principle schematic diagram of the barrier arrangement for perspective in another preferred embodiment.
Main element symbol description
100A the first barrier arrangement for perspective
100B the second barrier arrangement for perspective
10 first cameras
20 second cameras
30 barriers
40 target images
50 third cameras
60 the 4th cameras
101 first preset boundaries
102 second preset boundaries
103 first revise the boundary
201 third preset boundaries
202 the 4th preset boundaries
203 second revise the boundary
First visual field A1
Second visual field A2
B1 A Third Position
401 first reserved images
402 first shielded images
403 second reserved images
404 second shielded images
405 blending images
406 residual shielded images
The present invention that the following detailed description will be further explained with reference to the above drawings.
Specific embodiment
As depicted in figs. 1 and 2, barrier arrangement for perspective includes a photographic device (not indicating in figure) and a processing unit (not shown).Wherein, photographic device includes at least two groups camera, and the direction of at least two groups camera is different.Below The first camera 10 selected in barrier arrangement for perspective and second camera 20 are illustrated in embodiment.
The visual angle of first camera 10 and second camera 20 has overlapping region.First camera 10 and second camera 20 signal output end is connected to processing unit, which can be selected a microprocessor.Processing unit receives the first camera shooting First 10 and second camera 20 two groups of camera shooting information after, two groups of camera shooting information are merged, the panorama letter of a presumptive area is formed Breath.Panoramic information herein can be regarded as the adduction for the part that do not blocked by barrier 30 in two groups of camera shooting information, i.e. a width portion Divide or completely eliminate the image of shield portions.
It is elaborated below with reference to working principle of the barrier perspective method to barrier arrangement for perspective.
For example, the complete image of photographic device target image 40 to be acquired.However, due to the presence of barrier 30, first The visual field of camera 10 and second camera 20 by 30 partial occlusion of barrier, leads to the first camera 10, second camera 20 can not the complete target image 40 of independent acquisition complete image.At this point, due to the first camera 10, second camera 20 Visual angle it is different, and the visual angle of the first camera 10 and second camera 20 has an overlapping region, therefore the first camera 10 and the Two groups of camera shooting information of two cameras 20 can be merged mutually.
The first image is acquired by the first camera 10, the second image is acquired by second camera 20.
As shown in the upper left of Fig. 2, the theoretical visual field of the first camera 10 is preset by the first preset boundary 101 and second Boundary 102 is defined, and along both the first preset boundary 101 and the splicing of the second preset boundary 102.Due to depositing for barrier 30 The practical visual field of, the first camera 10 be the first preset boundary 101 and first revise the boundary 103 intermediate region;Therefore first Camera 10 can only collect target image 40 positioned at the first preset boundary 101 and the first amendment when acquiring target image 40 The first reserved image 401 between boundary 103, target image 40 be located at the first preset boundary 101 and first revise the boundary 103 it Outer part is then denoted as the first shielded image 402.
Similarly, as shown in the upper right portion of Fig. 2, the theoretical visual field of second camera 20 is by third preset boundary 201 and Four preset boundaries 202 define.Due to the presence of barrier 30, the practical visual field of second camera 20 is the 4th preset boundary 202 Revise the boundary with second 203 intermediate region can only collect target image 40 positioned at the 4th therefore when acquiring target image 40 Preset boundary 202 and second the second reserved image 403 for revising the boundary between 203, target image 40 are located at the 4th preset boundary 202 and second the part that revises the boundary except 203 be then denoted as the second shielded image 404.
First reserved image 401 and the second reserved image 403 cannot constitute the complete image of target image 40.First hides Gear image 402 and the second shielded image 404 form the blind area of the first camera 10, second camera 20.Actually answering In, it is desirable to reduce or eliminate blind area.
To achieve the purpose that reduce or eliminate blind area, can by along first revise the boundary 103 and second amendment side Boundary 203 splices, and the first reserved image 401 and the second reserved image 403 are superimposed.
Also that is, the stitching position of the first reserved image 401, the second reserved image 403 can be realized by the following method: according to The pre-set dimension of first camera 10, the visual angle of second camera 20 and the first image, the second image determines the first image, second The location information of image compares the location information of the first image, the second image, is referring to another image of calculating with a wherein image Relative to the angular transformation amount and translation transformation amount referring to image.
First reserved image 401, the second reserved image 403 stitching position can also realize by the following method: to One reserved image 401, the second reserved image 403 carry out feature extraction;It compares the feature extracted and carries out the first reserved image 401, the registration of the second reserved image 403.It compares the feature extracted and the step of carrying out image registration may particularly include: according to the The position of lap in one reserved image 401, the second reserved image 403 determines the reserved figure of the first reserved image 401, second As 403 transformation relation, the transformation of mapping relations between the reserved image 401 of to be spliced first, the second reserved image 403 is found out Model;The registration of multiple image is realized in the way of one or more registrations.The method of image registration can be used in the prior art Method realize that, because being the prior art of maturation, therefore details are not described herein.
It, can be by a processing unit, such as after the stitching position for determining the first reserved image 401, the second reserved image 403 One microprocessor splices the first reserved image 401, the second reserved image 403.First reserved image 401, second is reserved Image 403, which carries out splicing, to be realized by the method for Weighted Fusion.The blending algorithm that Pixel-level can be used in the Weighted Fusion is realized. The blending image 405 that first reserved image 401, the second reserved image 403 obtain after merging is as shown in the lower part of figure 2.Specific In implementation, the factors such as visual angle by the size of barrier 30, boundary and the first camera 10, second camera 20 are influenced, Blending image 405 may be the complete image of target image 40, it is also possible to there is residual shielded image 406, but finally can Achieve the purpose that blind area caused by reducing or eliminating barrier 30.
And be directed to the part of spliced residual shielded image 406, then it can handle in several ways;It is one of real Applying example is, to close on the reserved image to revise the boundary as map file, mirror image to be generated, to insert residual Occlusion Map using mirror image processing As 406;Another embodiment is to stick predetermined image, such as stick the trade mark defaulted, product image, other predetermined images, To fill up residual shielded image 406.
It, can be comprising steps of with wherein one the step of image 401, the second reserved image 403 are reserved in splicing first before It is reference that width, which reserves image, carries out size change to other reserved images in the first reserved image 401, the second reserved image 403 It changes.Further, it is also possible to reserve image as reference, to the first reserved image 401, the second reserved image using a wherein width as needed Other reserved images in 403 carry out resolution conversion, to realize subsequent Weighted Fusion.
It is only illustrated for using the first camera 10, second camera 20 in the present embodiment.It is easily understood that In other embodiments, it using more than two camera collection images and can also be merged as needed.
As shown in figure 3, there are multiple barrier arrangement for perspective in another application scene, with two barrier arrangement for perspective For, the third camera 50 of one first barrier arrangement for perspective A when shooting, can be in its theoretical visual field (i.e. first in Fig. 3 Visual field visual field A1) in there is one second barrier arrangement for perspective B, lead to third camera 50 there are one to block the visual field (i.e. in Fig. 3 The second visual field A2), if but user be not intended to the second barrier arrangement for perspective B appear in third camera 50 acquisition image In, then the parts of images in A Third Position B1 taken by the 4th camera 60 using the second barrier arrangement for perspective B The image of the second visual field part A2 is replaced, to obtain the image not blocked by the second barrier arrangement for perspective B.
In above-mentioned barrier perspective method and barrier arrangement for perspective, by being adopted from the camera lens of at least two different perspectivess After collection, at least two images blocked by barrier section, the part at least blocked in two images by barrier is deleted, it is raw Image is reserved at least two width, and splices at least two width and reserves image, then at least two width are reserved in image, not by barrier 30 The part blocked will be summed it up, to obtain the image that a width partially or completely eliminates shield portions, realize perspective barrier 30 purpose, and then to reduce or eliminate blind area caused by barrier 30.And this two camera lens can be deposited according to the difference of embodiment In same barrier arrangement for perspective, or it is stored in different barrier arrangement for perspective.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.

Claims (10)

1. a kind of barrier perspective method, comprising steps of
At least two images that scheduled barrier section blocks are acquired from least two different perspectivess;
The part blocked in at least two images by barrier is deleted, at least two width is generated and reserves image;
Determine that at least two width reserve the stitching position of image;
Splice at least two width and reserves image.
2. barrier perspective method as described in claim 1, it is characterised in that: adjacent at least two different perspectives Two visual angles have overlapping region.
3. barrier perspective method as described in claim 1, which is characterized in that determine that at least two width reserve the spelling of image The step of connecing position specifically includes:
At least two images are determined according to the pre-set dimension of at least two different perspectives and at least two images Location information;
The location information of adjacent image is compared, is referring to calculating angular transformation amount and translation transformation amount with a wherein image.
4. barrier perspective method as described in claim 1, which is characterized in that determine that at least two width reserve the spelling of image The step of connecing position specifically includes:
Feature extraction is carried out to adjacent reserved image;
It compares the feature extracted and carries out image registration.
5. barrier perspective method as described in claim 1, which is characterized in that described to compare the feature extracted and carry out image The step of registration, specifically includes:
According to the position of lap in adjacent image, the transformation relation of two images is determined, find out and reflect between image to be spliced Penetrate the transformation model of relationship;
The registration of multiple image is realized in the way of one or more registrations.
6. barrier perspective method as described in claim 1, which is characterized in that reserve image splicing at least two width Before step, further comprise the steps of:
Image is reserved as reference using a wherein width, other reserved images in image are reserved at least two width and carry out size change It changes.
7. barrier perspective method as described in claim 1 or 6, which is characterized in that splicing the reserved figure of at least two width Before the step of picture, further comprise the steps of:
Image is reserved as reference using a wherein width, other reserved images in image are reserved at least two width and carry out resolution ratio Transformation.
8. barrier perspective method as described in claim 1, it is characterised in that: splice the step that at least two width reserve image Suddenly it specifically includes:
To the corresponding region Weighted Fusion of adjacent reserved image, which is realized using the blending algorithm of Pixel-level.
9. a kind of barrier arrangement for perspective, including a photographic device and a processing unit, it is characterised in that: the photographic device packet At least two groups camera is included, the direction of at least two groups camera is different;The signal output end of at least two groups camera It is connected to the processing unit;
After the processing unit receives at least two groups camera shooting information of at least two groups camera, at least two groups are imaged Information fusion, forms the panoramic information of a presumptive area.
10. barrier arrangement for perspective as claimed in claim 9, it is characterised in that: in at least two groups camera, adjacent two The visual angle of group camera has overlapping region.
CN201711001278.0A 2017-10-24 2017-10-24 Barrier perspective method and barrier arrangement for perspective Pending CN109697745A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711001278.0A CN109697745A (en) 2017-10-24 2017-10-24 Barrier perspective method and barrier arrangement for perspective

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711001278.0A CN109697745A (en) 2017-10-24 2017-10-24 Barrier perspective method and barrier arrangement for perspective

Publications (1)

Publication Number Publication Date
CN109697745A true CN109697745A (en) 2019-04-30

Family

ID=66228046

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711001278.0A Pending CN109697745A (en) 2017-10-24 2017-10-24 Barrier perspective method and barrier arrangement for perspective

Country Status (1)

Country Link
CN (1) CN109697745A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111836103A (en) * 2019-12-31 2020-10-27 于刚 Anti-occlusion processing system based on data analysis

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5159455A (en) * 1990-03-05 1992-10-27 General Imaging Corporation Multisensor high-resolution camera
US20080118155A1 (en) * 2006-11-06 2008-05-22 Sony Corporation Image processing apparatus, camera apparatus, image processing method, and program
US20100118116A1 (en) * 2007-06-08 2010-05-13 Wojciech Nowak Tomasz Method of and apparatus for producing a multi-viewpoint panorama
CN103810686A (en) * 2014-02-27 2014-05-21 苏州大学 Seamless splicing panorama assisting driving system and method
CN104021388A (en) * 2014-05-14 2014-09-03 西安理工大学 Reversing obstacle automatic detection and early warning method based on binocular vision
CN104580882A (en) * 2014-11-03 2015-04-29 宇龙计算机通信科技(深圳)有限公司 Photographing method and device
JP2017117244A (en) * 2015-12-24 2017-06-29 Kddi株式会社 Image analysis device for detecting a person displayed in photographic image of camera

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5159455A (en) * 1990-03-05 1992-10-27 General Imaging Corporation Multisensor high-resolution camera
US20080118155A1 (en) * 2006-11-06 2008-05-22 Sony Corporation Image processing apparatus, camera apparatus, image processing method, and program
US20100118116A1 (en) * 2007-06-08 2010-05-13 Wojciech Nowak Tomasz Method of and apparatus for producing a multi-viewpoint panorama
CN103810686A (en) * 2014-02-27 2014-05-21 苏州大学 Seamless splicing panorama assisting driving system and method
CN104021388A (en) * 2014-05-14 2014-09-03 西安理工大学 Reversing obstacle automatic detection and early warning method based on binocular vision
CN104580882A (en) * 2014-11-03 2015-04-29 宇龙计算机通信科技(深圳)有限公司 Photographing method and device
JP2017117244A (en) * 2015-12-24 2017-06-29 Kddi株式会社 Image analysis device for detecting a person displayed in photographic image of camera

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
韩晓微等: "《实景图像拼接及其漫游控制技术》", 普通高等教育电子信息类"十三五"课改规划教材, pages: 55 - 107 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111836103A (en) * 2019-12-31 2020-10-27 于刚 Anti-occlusion processing system based on data analysis
CN111836103B (en) * 2019-12-31 2022-04-05 北京无忧乐道科技有限公司 Anti-occlusion processing system based on data analysis

Similar Documents

Publication Publication Date Title
US20190098282A1 (en) Stereo imaging system with automatic disparity adjustment for displaying close range objects
JP5178553B2 (en) Imaging device
US6628820B2 (en) Image forming method and apparatus
US5602584A (en) Apparatus for producing a panoramic image using a plurality of optical systems
EP2340649B1 (en) Three-dimensional display device and method as well as program
CN106056534A (en) Obstruction perspective method and device based on smart glasses
US20070165129A1 (en) Method of and apparatus for selecting a stereoscopic pair of images
CN106412426A (en) Omni-focus photographing apparatus and method
CN109165629B (en) It is multifocal away from visual barrier cognitive method, device, equipment and storage medium
DE102013220669A1 (en) Dynamic rearview indicator features
DE112012001022T5 (en) Alignment control in a head-worn augmented reality device
JPH1098646A (en) Object extraction system
WO2012152475A1 (en) Method and device for calibrating a projection device of a vehicle
US11528453B2 (en) Sensor fusion based perceptually enhanced surround view
CN106851107A (en) Switch control method, control device and the electronic installation of camera assisted drawing
US20070273757A1 (en) Panoramic viewing system especially in combat vehicles
JP2005020559A (en) Three-dimensional video presentation apparatus
CN109360150A (en) A kind of joining method and device of the panorama depth map based on depth camera
WO2018134897A1 (en) Position and posture detection device, ar display device, position and posture detection method, and ar display method
CN101540822A (en) Device and method for high-resolution large-viewing-field aerial image forming
JP2006060425A (en) Image generating method and apparatus thereof
CN112673621A (en) Automatic focusing method, device, holder, equipment and storage medium
JP2006119843A (en) Image forming method, and apparatus thereof
CN109697745A (en) Barrier perspective method and barrier arrangement for perspective
TWI653625B (en) Method and device for seeing through obstacles

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
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20190430