WO2016155110A1 - 图像透视畸变校正的方法及*** - Google Patents

图像透视畸变校正的方法及*** Download PDF

Info

Publication number
WO2016155110A1
WO2016155110A1 PCT/CN2015/080314 CN2015080314W WO2016155110A1 WO 2016155110 A1 WO2016155110 A1 WO 2016155110A1 CN 2015080314 W CN2015080314 W CN 2015080314W WO 2016155110 A1 WO2016155110 A1 WO 2016155110A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
vector information
perspective distortion
distortion correction
dimensional vector
Prior art date
Application number
PCT/CN2015/080314
Other languages
English (en)
French (fr)
Inventor
高秀文
Original Assignee
宇龙计算机通信科技(深圳)有限公司
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 宇龙计算机通信科技(深圳)有限公司 filed Critical 宇龙计算机通信科技(深圳)有限公司
Publication of WO2016155110A1 publication Critical patent/WO2016155110A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/80—Geometric correction

Definitions

  • the present invention relates to the field of image capturing technology, and in particular, to a method and system for image perspective distortion correction.
  • the optical system of the camera terminal such as a camera and a mobile phone
  • there is perspective distortion and there is an optical distortion error between the actual imaging and the ideal imaging of the object on the imaging surface of the imaging terminal.
  • there is a technique of performing lens perspective distortion correction based on physical distance and direction which takes several images by lens focusing, then analyzes the image to determine the focus calculation distance, and then uses the distance parameter to perform perspective distortion correction on the image.
  • the correction accuracy is low, resulting in poor correction effect; and each time a plurality of photos are taken, a single time adjustment, exposure, etc. are required for a long time and the speed is slow.
  • the prior art can only correct the perspective distortion of a two-dimensional (2D) image, and the perspective distortion of a three-dimensional (3D) image is at a loss.
  • an object of the present invention is to provide a method and system for correcting image perspective distortion, which can perform perspective distortion correction on a three-dimensional image and can greatly improve the speed of perspective distortion correction.
  • the present invention provides a method for image perspective distortion correction, which is applied to a package.
  • a camera terminal comprising two cameras, the method comprising:
  • the step of respectively generating the first three-dimensional vector information corresponding to the first image and the second three-dimensional vector information corresponding to the second image includes:
  • the step of respectively generating the first three-dimensional vector information corresponding to the first image and the second three-dimensional vector information corresponding to the second image includes:
  • a coincidence region of the first image and the second image is analyzed
  • the first image and the second image are performed by a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information.
  • the steps of perspective distortion correction include:
  • the first image and the second image are obtained by a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information
  • the steps include:
  • the image display instruction is a two-dimensional image display instruction, intercepting the corrected overlapping area in the first image or the second image for display;
  • the corrected first image and the second image are three-dimensionally modulated and displayed.
  • the present invention also provides a system for image perspective distortion correction applied to a photographing terminal including two cameras, the system comprising:
  • An image acquisition module is configured to simultaneously capture a subject by using the first camera and the second camera to acquire a corresponding first image and a second image;
  • An information generating module configured to respectively generate first three-dimensional vector information corresponding to the first image and second three-dimensional vector information corresponding to the second image;
  • an image correction module configured to perform perspective distortion correction on the first image and the second image by using a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information.
  • the information generating module includes:
  • a first generation submodule configured to generate the first three-dimensional vector information of each pixel in the first image
  • a second generation submodule configured to generate the second three-dimensional vector information of each pixel in the second image.
  • the information generating module includes:
  • a region analysis submodule configured to analyze a coincident region of the first image and the second image
  • the first generation sub-module is configured to calculate a distance of an original image point corresponding to each of the pixel points in the overlapping area from the first camera to obtain a first distance, and according to each Generating the first three-dimensional vector information of the first image and the first two-dimensional coordinates of the first image;
  • the second generation sub-module is configured to calculate a distance of an original image point corresponding to each of the pixel points in the overlapping area from the second camera to obtain a second distance, and according to each Generating a second two-dimensional coordinate of the pixel and the second distance to generate the second three-dimensional image of the second image Vector information.
  • the image correction module includes:
  • a first calibration submodule configured to invoke a predetermined first perspective distortion correction parameter according to the first three-dimensional vector information, and perform perspective distortion correction on the coincident region of the first image by using the perspective distortion correction algorithm;
  • a second calibration submodule configured to invoke a predetermined second perspective distortion correction parameter according to the second three-dimensional vector information, and perform perspective distortion correction on the overlapping region of the second image by the perspective distortion correction algorithm.
  • the system according to the invention further comprises:
  • the instruction receiving module is configured to determine, after receiving the image display instruction, the image display instruction, after performing the perspective distortion correction on the first image and the second image;
  • a first display module configured to: when the image display instruction is a two-dimensional image display instruction, intercept the corrected overlapping area in the first image or the second image for display;
  • a second display module configured to perform three-dimensional modulation and display on the corrected first image and the second image if the image display instruction is a three-dimensional image display instruction.
  • the photographing terminal of the present invention captures two images of the object at a time through two cameras, and respectively generates three-dimensional vector information of the two images
  • the prior art needs to be photographed multiple times with different distances.
  • the present invention can achieve perspective distortion correction for a three-dimensional image, and can greatly improve the speed of perspective distortion correction.
  • the present invention can capture three-dimensional vector information of each pixel in the captured image by using two cameras, and perform perspective distortion correction of the three-dimensional image according to the three-dimensional vector information of each pixel point, and several sheets are compared compared with the prior art.
  • the accuracy of the local wide-range correction is obtained by several depth values obtained by the photograph, and the precision of the perspective distortion correction is greatly improved by the present invention.
  • FIG. 1 is a schematic structural view of a system for correcting image distortion of the present invention
  • FIG. 2 is a schematic structural view of a system for optimizing image perspective distortion correction of the present invention
  • FIG. 3 is a flow chart of a method for correcting perspective distortion of an image of the present invention
  • FIG. 4 is a flow chart of a method for correcting image perspective distortion in a first embodiment of the present invention
  • Figure 5 is a schematic diagram showing the implementation of an image perspective distortion correction method in a second embodiment of the present invention.
  • the system 100 includes an image acquisition module 10, an information generation module 20, and an image correction module 30, wherein:
  • the image acquisition module 10 is configured to simultaneously capture a subject by using the first camera and the second camera to acquire a corresponding first image and second image.
  • the subject may be any one or more of a person, an animal, a plant, a building, a mountain, a water, a sky, and the like.
  • the first camera and the second camera are two cameras on the same side of the shooting terminal, and may be left and right cameras or upper and lower cameras.
  • the two cameras can simultaneously capture and the captured content is basically the same, that is, the first image and the second image.
  • the image content of the image is basically the same, and there are many overlapping areas.
  • two cameras have been widely used in shooting cameras such as cameras, mobile phones, and tablets.
  • the information generating module 20 is configured to respectively generate first three-dimensional vector information corresponding to the first image and second three-dimensional vector information corresponding to the second image.
  • the first three-dimensional vector information may be generated according to the first image
  • the first three-dimensional vector information may constitute a first three-dimensional image
  • the second three-dimensional vector information may be generated according to the second image, where the second three-dimensional vector information may constitute the second three-dimensional image image.
  • the three-dimensional vector information in the present invention refers to the planar two-dimensional coordinates of the image and the vector information of the distance between the original image point and the camera corresponding to each pixel point in the image.
  • the distance is preferably a plane two-dimensional coordinate of the image and a distance between the original image point corresponding to each pixel point in the image and the camera plane.
  • the distance may also be a planar two-dimensional coordinate of the image and each pixel in the image.
  • the original image point is a point of an actual object corresponding to a pixel point in the image.
  • the image correction module 30 is configured to perform perspective distortion correction on the first image and the second image by using a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information.
  • At least one perspective distortion correction algorithm may be pre-stored in the photographing terminal, for example based on an object
  • the lens perspective distortion correction algorithm is performed on the distance and the direction. Since the perspective distortion correction algorithm is prior art, it will not be described herein. Specifically, the first image is subjected to perspective distortion correction according to the perspective distortion correction algorithm and the first three-dimensional vector information; and the second image is subjected to perspective distortion correction according to the perspective distortion correction algorithm and the second three-dimensional vector information.
  • the photographing terminal of the present invention captures two images of the object at a time through two cameras, and respectively generates three-dimensional vector information of the two images, the prior art needs to be photographed multiple times with different distances.
  • FIG. 2 is a schematic structural diagram of a preferred image perspective distortion correction system of the present invention, applied to a photographing terminal including two cameras, the photographing terminal may be a camera, a mobile phone, a tablet computer, etc., and the system 100 includes at least an image acquisition module. 10.
  • the image acquisition module 10 is configured to simultaneously capture a subject by using the first camera and the second camera to acquire a corresponding first image and second image.
  • the information generating module 20 is configured to respectively generate first three-dimensional vector information corresponding to the first image and second three-dimensional vector information corresponding to the second image.
  • the information generating module 20 includes:
  • the first generation sub-module 21 is configured to generate first three-dimensional vector information of each pixel in the first image.
  • the second generation sub-module 22 is configured to generate second three-dimensional vector information of each pixel in the second image.
  • the three-dimensional vector information is acquired by one camera at a time, which greatly shortens the time required for the prior art to take multiple photos to obtain depth information by multiple times.
  • the accuracy of the three-dimensional vector information correction based on the pixel is ten million times the precision of the local large-scale correction by several depth values obtained by several photos in the prior art, and the speed and accuracy of the perspective distortion correction can be greatly improved.
  • the information generating module 30 may further include:
  • the area analysis sub-module 23 is configured to analyze the overlapping area of the first image and the second image.
  • the analysis of the coincident region involves image recognition technology based on RGB (Red, Green, Blue, Red, green and blue) values, brightness, grayscale and other pixel features find the boundary between the similar point and the dissimilar point, and the overlapping area of the first image and the second image can be analyzed.
  • RGB Red, Green, Blue, Red, green and blue
  • the first generation sub-module 21 is configured to calculate a distance of the original image point corresponding to each pixel point of the first image in the coincidence region from the first camera to obtain a first distance, and according to the first two-dimensional coordinates of each pixel point And the first distance, generating first three-dimensional vector information of the first image, and the first three-dimensional vector information may be represented by (x, y, z). That is, the first three-dimensional vector information in the present invention refers to the planar two-dimensional coordinates of the first image and the vector information of the distance between the original image point corresponding to each pixel point in the first image and the first camera.
  • the second generation sub-module 22 is configured to calculate a distance of the original image point corresponding to each pixel point of the second image in the coincidence region from the second camera to obtain a second distance, and according to the second two-dimensional coordinate of each pixel point And the second distance, generating second three-dimensional vector information of the second image, and the second three-dimensional vector information may be represented by (x, y, z). That is, the second three-dimensional vector information in the present invention refers to the planar two-dimensional coordinates of the second image and the vector information of the distance between the original image point and the second camera corresponding to each pixel point in the second image.
  • the image correction module 30 is configured to perform perspective distortion correction on the first image and the second image by using a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information.
  • the image correction module 30 includes:
  • the first syndrome module 31 is configured to invoke a predetermined first perspective distortion correction parameter according to the first three-dimensional vector information, and perform perspective distortion correction on the coincident region of the first image by the perspective distortion correction algorithm.
  • the second calibration sub-module 32 is configured to invoke a predetermined second perspective distortion correction parameter according to the second three-dimensional vector information, and perform perspective distortion correction on the coincident region of the second image by the perspective distortion correction algorithm.
  • the second perspective distortion correction parameter and the first perspective distortion correction parameter may be the same or different.
  • the image perspective distortion correction system 100 can further include:
  • the instruction receiving module 40 is configured to determine the type of the image display instruction after receiving the image display instruction after performing the perspective distortion correction on the first image and the second image.
  • the type of the image display instruction may be a two-dimensional image display instruction or a three-dimensional image display instruction.
  • the first display module 50 is configured to intercept the coincident area in the corrected first image or the second image for display if the image display instruction is a two-dimensional image display instruction.
  • the second display module 60 is configured to perform three-dimensional modulation and display on the corrected first image and the second image when the image display instruction is a three-dimensional image display instruction.
  • the corrected first image and the second image are simultaneously displayed to form a three-dimensional image.
  • the invention describes a technique for capturing three-dimensional vector information of each pixel point in a captured image by two cameras, and calling the corresponding lens perspective distortion parameter to perform three-dimensional image distortion correction technology, thereby greatly improving the speed and precision of the perspective distortion correction, and making up for the present
  • the technique cannot perform perspective distortion correction on a three-dimensional image. You can also use the corrected two images for 3D display, and the user experience is better.
  • FIG. 3 is a flow chart of a method for image perspective distortion correction of the present invention applied to a photographing terminal including two cameras, the method being implemented by a system 100 for perspective distortion correction as shown in FIG. 1 or FIG. 2, the method Includes:
  • Step S301 the subject is simultaneously photographed by the first camera and the second camera, and the corresponding first image and second image are acquired.
  • the subject may be any one or more of a person, an animal, a plant, a building, a mountain, a water, a sky, and the like.
  • the first camera and the second camera are two cameras on the same side of the shooting terminal, and may be left and right cameras or upper and lower cameras.
  • the two cameras can simultaneously capture and the captured content is basically the same, that is, the first image and the second image.
  • the image content of the image is basically the same, and there are many overlapping areas.
  • Step S302 respectively generating first three-dimensional vector information corresponding to the first image and second three-dimensional vector information corresponding to the second image.
  • the first three-dimensional vector information may be generated according to the first image, the first three-dimensional vector information may constitute a first three-dimensional image; and the second three-dimensional vector information may be generated according to the second image, where the second three-dimensional vector information may constitute the second three-dimensional image image.
  • first three-dimensional vector information of each pixel in the first image is generated, and second three-dimensional vector information of each pixel in the second image is generated.
  • the accuracy of the three-dimensional vector information correction based on the pixel is ten million times the precision of the local large-scale correction by several depth values obtained by several photos in the prior art, and the speed and accuracy of the perspective distortion correction can be greatly improved.
  • the three-dimensional vector information in the present invention refers to the flatness of the image.
  • the distance is preferably a plane two-dimensional coordinate of the image and a distance between the original image point corresponding to each pixel point in the image and the camera plane.
  • the distance may also be a planar two-dimensional coordinate of the image and each pixel in the image.
  • the original image point is a point of an actual object corresponding to a pixel point in the image.
  • Step S303 performing perspective distortion correction on the first image and the second image by using a predetermined perspective distortion correction algorithm according to the obtained first three-dimensional vector information and the second three-dimensional vector information.
  • At least one perspective distortion correction algorithm may be pre-stored in the photographing terminal, for example, a lens perspective distortion correction algorithm based on the object distance and direction. Since the perspective distortion correction algorithm is prior art, it will not be described herein. Specifically, the first image is subjected to perspective distortion correction according to the perspective distortion correction algorithm and the first three-dimensional vector information; and the second image is subjected to perspective distortion correction according to the perspective distortion correction algorithm and the second three-dimensional vector information.
  • FIG. 4 is a flow chart of a method for correcting image perspective distortion in a first embodiment of the present invention, applied to a photographing terminal including two cameras, which can be realized by a system 100 for perspective distortion correction as shown in FIG.
  • the methods include:
  • Step S401 the subject is simultaneously photographed by the first camera and the second camera, and the corresponding first image and second image are acquired.
  • Step S402 analyzing a coincident region of the first image and the second image.
  • the analysis of the coincidence region involves an image recognition technology, and the coincidence region of the first image and the second image can be analyzed by finding a boundary between the similar point and the dissimilarity point based on pixel feature such as RGB value, brightness, and gray scale.
  • Step S403 calculating a distance of the original image point corresponding to each pixel point of the first image in the coincidence region from the first camera, obtaining a first distance, and generating according to the first two-dimensional coordinates and the first distance of each pixel point.
  • the first three-dimensional vector information of the first image is
  • the first three-dimensional vector information in the present invention refers to the planar two-dimensional coordinates of the first image and the vector information of the distance between the corresponding pixel point and the first camera in the first image, and the first three-dimensional vector information can be used (x, y , z) to express.
  • Step S404 invoking a predetermined first perspective distortion correction parameter according to the first three-dimensional vector information, and performing perspective distortion correction on the coincident region of the first image by a perspective distortion correction algorithm.
  • Step S405 calculating an original image point distance corresponding to each pixel point of the second image in the coincidence area
  • the distance between the two cameras obtains a second distance, and generates second three-dimensional vector information of the second image according to the second two-dimensional coordinates and the second distance of each pixel.
  • the second three-dimensional vector information in the present invention refers to the planar two-dimensional coordinates of the second image and the vector information of the distance between the original image point and the second camera corresponding to each pixel point in the second image, and the second three-dimensional vector information may Expressed by (x, y, z).
  • Step S406 invoking a predetermined second perspective distortion correction parameter according to the second three-dimensional vector information, and performing perspective distortion correction on the coincident region of the second image by the perspective distortion correction algorithm.
  • the second perspective distortion correction parameter and the first perspective distortion correction parameter may be the same or different.
  • Step S407 an image display instruction is received.
  • the type of the image display instruction may be a two-dimensional image display instruction or a three-dimensional image display instruction.
  • step S408 the type of the image display instruction is determined. If the image display instruction is a two-dimensional image display instruction, step S409 is performed, and if the image display instruction is a three-dimensional image display instruction, step S410 is performed.
  • Step S409 if the image display instruction is a two-dimensional image display instruction, the coincident region in the corrected first image or the second image is cut out for display.
  • Step S410 If the image display instruction is a three-dimensional image display instruction, the corrected first image and the second image are three-dimensionally modulated and displayed.
  • the corrected first image and the second image are simultaneously displayed to form a three-dimensional image.
  • FIG. 5 is a schematic diagram of an implementation of an image perspective distortion correction method according to a second embodiment of the present invention, which is applied to a photographing terminal including a left and right camera, and mainly includes:
  • the subject is photographed simultaneously by the first camera and the second camera, and the corresponding first image and second image are acquired.
  • the left and right cameras simultaneously take pictures of the subject, and obtain two images of the left and right and save, that is, the image L (ie, the first image) and the image R (ie, the second image);
  • the first coordinate of the image L, the second coordinate of the image R, and the third coordinate of the coincident image formed by the overlapping portions of the two images are three coordinate systems.
  • the first coordinate of the image L, the second coordinate of the image R is a two-dimensional coordinate system
  • the third coordinate of the coincident image is a three-dimensional coordinate system.
  • the X-axis origin in the third coordinate (x, y, z) is the abscissa x1 of the first column coincident pixel point in the first coordinate of the image L
  • the Y-axis origin in the third coordinate (x, y, z) To coincide the ordinate ym of the pixel of the bottom row in the image.
  • the coordinate origin refers to the last pixel of the lower left corner of each image (image L, image R, or coincident image).
  • image L image R
  • image R image R
  • the coordinate origin of the third coordinate is transformed based on the coordinate origin of the first coordinate or the second coordinate, and the conversion is used to x1 or x2.
  • the perspective distortion correction coefficient table provided by the lens manufacturer is called to perform perspective distortion correction on the coincident region (the oblique line portion in the figure) in the two images by the calibration algorithm (the non-coincident region correction value defaults to 0).
  • the distortion coefficient M related to the three-dimensional vector information (x, y, z) in the correction process corresponds to the pixel point of the left camera captured image L
  • M corresponds to the pixel point of the right camera captured image R
  • only the corrected left and right images L are saved.
  • J , R J delete other temporary data or files.
  • R J can perform three-dimensional modulation and simultaneous playback.
  • the photographing terminal of the present invention acquires two images of the object at a time through two cameras and generates three-dimensional vector information of the two images, which can greatly shorten the prior art and require different distances.
  • the sub-focus captures the time at which the plurality of images acquire depth information; then the perspective distortion correction algorithm is invoked to perform perspective distortion correction on the three-dimensional image.
  • the present invention can achieve perspective distortion correction for a three-dimensional image, and can greatly improve the speed of perspective distortion correction.
  • the present invention can capture three-dimensional vector information of each pixel in the captured image by using two cameras, and perform perspective distortion correction of the three-dimensional image according to the three-dimensional vector information of each pixel point, and several sheets are compared compared with the prior art.
  • the accuracy of the local wide-range correction is obtained by several depth values obtained by the photograph, and the precision of the perspective distortion correction is greatly improved by the present invention.

Landscapes

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

Abstract

本发明适用于图像技术领域，提供了一种图像透视畸变校正的方法及***，应用于包括两个摄像头的拍摄终端，所述方法包括有：通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像；分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息；根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正。借此，本发明能够实现对三维图像进行透视畸变校正，并且可大大提高透视畸变校正的速度。

Description

图像透视畸变校正的方法及***
本申请要求于2015年03月31日提交中国专利局，申请号为CN 201510149800.4、发明名称为“图像透视畸变校正的方法及***”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及图像拍摄技术领域，尤其涉及一种图像透视畸变校正的方法及***。
背景技术
由于相机、手机等拍摄终端光学***并不是精确地按理想化的小孔成像的原理工作，存在有透视畸变，物体在拍摄终端的成像面上的实际成像与理想成像之间存在有光学畸变误差。现有技术中已存在基于物理距离与方向进行镜头透视畸变校正的技术，其通过镜头调焦拍摄几张图像，然后分析图像判断焦点计算距离，再利用距离参数对图像进行透视畸变校正。现有技术由于不可能拍摄无数张照片来计算距离，所以校正精度很低导致校正效果不好；而且拍摄多张照片每次都需单次调焦、曝光等所需时间很长从而速度慢。另外，现有技术只能校正二维(Two Dimension，2D)图像的透视畸变，对三维(Three Dimension，3D)图像的透视畸变则束手无策。
综上可知，现有技术在实际使用上显然存在不便与缺陷，所以有必要加以改进。
发明内容
针对上述的缺陷，本发明的目的在于提供一种图像透视畸变校正的方法及***，其能够实现对三维图像进行透视畸变校正，并且可大大提高透视畸变校正的速度。
为了实现上述目的，本发明提供一种图像透视畸变校正的方法，应用于包 括两个摄像头的拍摄终端，所述方法包括有：
通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像；
分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息；
根据得到所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正。
根据本发明所述的方法，所述分别生成第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息的步骤包括：
生成所述第一图像中各像素点的所述第一三维矢量信息；
生成所述第二图像中各像素点的所述第二三维矢量信息。
根据本发明所述的方法，所述分别生成第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息的步骤包括：
分析出所述第一图像和所述第二图像的重合区域；
计算所述第一图像在所述重合区域中每个所述像素点对应的原像点距所述第一摄像头的距离，得到第一距离，并根据每个所述像素点的第一二维坐标和所述第一距离，生成所述第一图像的所述第一三维矢量信息；
计算所述第二图像在所述重合区域中每个所述像素点对应的原像点距所述第二摄像头的距离，得到第二距离，并根据每个所述像素点的第二二维坐标和所述第二距离，生成所述第二图像的所述第二三维矢量信息。
根据本发明所述的方法，所述根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正的步骤包括：
根据所述第一三维矢量信息调用预定的第一透视畸变校正参数，并通过所述透视畸变校正算法对所述第一图像的所述重合区域进行透视畸变校正；
根据所述第二三维矢量信息调用预定的第二透视畸变校正参数，并通过所述透视畸变校正算法对所述第二图像的所述重合区域进行透视畸变校正。
根据本发明所述的方法，所述根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二 图像进行透视畸变校正的步骤之后包括：
若接收到图像展示指令，判断所述图像展示指令的类型；
若所述图像展示指令为二维图像展示指令，则截取出校正后的所述第一图像或所述第二图像中的所述重合区域进行展示；
若所述图像展示指令为三维图像展示指令，则对校正后的所述第一图像和所述第二图像进行三维调制和展示。
本发明还提供一种图像透视畸变校正的***，应用于包括两个摄像头的拍摄终端，所述***包括有：
图像获取模块，用于通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像；
信息生成模块，用于分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息；
图像校正模块，用于根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正。
根据本发明所述的***，所述信息生成模块包括：
第一生成子模块，用于生成所述第一图像中各像素点的所述第一三维矢量信息；
第二生成子模块，用于生成所述第二图像中各像素点的所述第二三维矢量信息。
根据本发明所述的***，所述信息生成模块包括：
区域分析子模块，用于分析出所述第一图像和所述第二图像的重合区域；
所述第一生成子模块，用于计算所述第一图像在所述重合区域中每个所述像素点对应的原像点距所述第一摄像头的距离，得到第一距离，并根据每个所述像素点的第一二维坐标和所述第一距离，生成所述第一图像的所述第一三维矢量信息；
所述第二生成子模块，用于计算所述第二图像在所述重合区域中每个所述像素点对应的原像点距所述第二摄像头的距离，得到第二距离，并根据每个所述像素点的第二二维坐标和所述第二距离，生成所述第二图像的所述第二三维 矢量信息。
根据本发明所述的***，所述图像校正模块包括：
第一校正子模块，用于根据所述第一三维矢量信息调用预定的第一透视畸变校正参数，并通过所述透视畸变校正算法对所述第一图像的所述重合区域进行透视畸变校正；
第二校正子模块，用于根据所述第二三维矢量信息调用预定的第二透视畸变校正参数，并通过所述透视畸变校正算法对所述第二图像的所述重合区域进行透视畸变校正。
根据本发明所述的***，还包括：
指令接收模块，用于对所述第一图像和所述第二图像进行透视畸变校正之后，若接收到图像展示指令，判断所述图像展示指令的类型；
第一展示模块，用于若所述图像展示指令为二维图像展示指令时，截取出校正后的所述第一图像或所述第二图像中的所述重合区域进行展示；
第二展示模块，用于若所述图像展示指令为三维图像展示指令时，对校正后的所述第一图像和所述第二图像进行三维调制和展示。
本发明拍摄终端在拍摄时，通过两个摄像头一次性获取被摄物体的两个图像，并分别生成两个图像的三维矢量信息，可大大缩短现有技术需不同距离分别多次对焦拍摄多个图像获取深度信息的时间；然后调用透视畸变校正算法对三维图像进行透视畸变校正。借此，本发明能够实现对三维图像进行透视畸变校正，并且可大大提高透视畸变校正的速度。优选的是，本发明可通过两个摄像头捕获被摄图像中各像素点的三维矢量信息，并根据各像素点的三维矢量信息进行三维图像的透视畸变校正，相比于现有技术通过几张照片获取的几个深度值进行局部大范围校正的精度，本发明大大提高了透视畸变校正的精度。
附图说明
图1是本发明图像透视畸变校正的***的结构示意图；
图2是本发明优选图像透视畸变校正的***的结构示意图；
图3是本发明图像透视畸变校正的方法流程图；
图4是本发明第一实施例中图像透视畸变校正的方法流程图；
图5是本发明第二实施例中图像透视畸变校正方法的实现原理图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
图1是本发明图像透视畸变校正的***的结构示意图，应用于包括两个摄像头的拍摄终端，所述***100包括有图像获取模块10、信息生成模块20以及图像校正模块30，其中：
图像获取模块10，用于通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像。所述被摄物体可以是人物、动物、植物、建筑物、山、水、天空等任意一种或多种。所述第一摄像头和第二摄像头为拍摄终端在同一面的两个摄像头，可以是左右摄像头或上下摄像头，所述两个摄像头可同时拍摄且拍摄的内容基本一致，即第一图像和第二图像的图像内容基本一致，存在较多的重合区域。目前两个摄像头已在相机、手机、平板电脑等拍摄终端上广泛应用。
信息生成模块20，用于分别生成第一图像对应的第一三维矢量信息和第二图像对应的第二三维矢量信息。这里，根据第一图像可以生成第一三维矢量信息，该第一三维矢量信息可以构成第一三维图像；根据第二图像可以生成第二三维矢量信息，该第二三维矢量信息可以构成第二三维图像。优选的是，本发明中的三维矢量信息是指图像的平面二维坐标以及图像中各像素点对应的原像点与摄像头之间距离的矢量信息。所述距离优选为图像的平面二维坐标以及图像中各应像素点对应的原像点与摄像头平面之间距离，当然所述距离也可以是图像的平面二维坐标以及图像中各应像素点对应的原像点与摄像头的中心点之间距离。所述原像点是图像中像素点对应的实际物体的点。
图像校正模块30，用于根据得到的第一三维矢量信息和第二三维矢量信息，通过预定的透视畸变校正算法，对第一图像和第二图像进行透视畸变校正。拍摄终端中可预存有至少一种透视畸变校正算法，例如基于物体 距离与方向进行镜头透视畸变校正算法，由于所述透视畸变校正算法为现有技术，因此在此不再赘述。具体而言，根据透视畸变校正算法和第一三维矢量信息，对第一图像进行透视畸变校正；以及根据透视畸变校正算法和第二三维矢量信息，对第二图像进行透视畸变校正。
本发明拍摄终端在拍摄时，通过两个摄像头一次性获取被摄物体的两个图像，并分别生成两个图像的三维矢量信息，可大大缩短现有技术需不同距离分别多次对焦拍摄多个图像获取深度信息的时间；然后调用透视畸变校正算法对三维图像进行透视畸变校正，从而实现对三维图像进行透视畸变校正。
图2是本发明优选图像透视畸变校正的***的结构示意图，应用于包括两个摄像头的拍摄终端，所述拍摄终端可以是相机、手机、平板电脑等，所述***100至少包括有图像获取模块10、信息生成模块20以及图像校正模块30，其中：
所述图像获取模块10，用于通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像。
所述信息生成模块20，用于分别生成第一图像对应的第一三维矢量信息和第二图像对应的第二三维矢量信息。优选的是，所述信息生成模块20包括：
第一生成子模块21，用于生成第一图像中各像素点的第一三维矢量信息。
第二生成子模块22，用于生成第二图像中各像素点的第二三维矢量信息。
通过两个摄像头一次拍摄获取三维矢量信息，大大缩短现有技术需不同距离分别多次对焦拍摄多张照片获取深度信息的时间。基于像素点的三维矢量信息校正的精度是现有技术通过几张照片获取的几个深度值进行局部大范围校正的精度的千万倍，可大大提高透视畸变校正的速度与精度。
更好的是，所述信息生成模块30还可包括：
区域分析子模块23，用于分析出第一图像和第二图像的重合区域。所述重合区域的分析涉及到图像识别技术，基于RGB(Red，Green，Blue， 红绿蓝)值，亮度，灰度等像素点特征找到相似点与不相似点分界线，即可分析出第一图像和第二图像的重合区域。
第一生成子模块21，用于计算第一图像在重合区域中每个像素点对应的原像点距第一摄像头的距离，得到第一距离，并根据每个像素点的第一二维坐标和第一距离，生成第一图像的第一三维矢量信息，第一三维矢量信息可以用(x，y，z)来表示。即本发明中的第一三维矢量信息是指第一图像的平面二维坐标以及第一图像中各像素点对应的原像点与第一摄像头之间距离的矢量信息。
第二生成子模块22，用于计算第二图像在重合区域中每个像素点对应的原像点距第二摄像头的距离，得到第二距离，并根据每个像素点的第二二维坐标和第二距离，生成第二图像的第二三维矢量信息，第二三维矢量信息可以用(x，y，z)来表示。即本发明中的第二三维矢量信息是指第二图像的平面二维坐标以及第二图像中中各像素点对应的原像点与第二摄像头之间距离的矢量信息。
所述图像校正模块30，用于根据得到的第一三维矢量信息和第二三维矢量信息，通过预定的透视畸变校正算法，对第一图像和第二图像进行透视畸变校正。
优选的是，所述图像校正模块30包括：
第一校正子模块31，用于根据第一三维矢量信息调用预定的第一透视畸变校正参数，并通过透视畸变校正算法对第一图像的重合区域进行透视畸变校正。
第二校正子模块32，用于根据第二三维矢量信息调用预定的第二透视畸变校正参数，并通过透视畸变校正算法对第二图像的重合区域进行透视畸变校正。第二透视畸变校正参数与第一透视畸变校正参数可以相同或不同。
更好的是，所述图像透视畸变校正的***100还可包括：
指令接收模块40，用于对第一图像和第二图像进行透视畸变校正之后，若接收到图像展示指令，判断所述图像展示指令的类型。所述图像展示指令的类型可以是二维图像展示指令或者三维图像展示指令。
第一展示模块50，用于若图像展示指令为二维图像展示指令时，截取出校正后的第一图像或第二图像中的重合区域进行展示。
第二展示模块60，用于若图像展示指令为三维图像展示指令时，对校正后的第一图像和第二图像进行三维调制和展示。这里，校正后的第一图像和第二图像同时展示可构成三维图像。
本发明描述了通过两个摄像头捕获被摄图像中各像素点的三维矢量信息，并调用相应镜头透视畸变参数进行三维图像畸变校正的技术，大大提高透视畸变校正的速度与精度，并且弥补了现有技术不能对三维图像进行透视畸变校正的缺陷。还可以用校正后的两幅图像进行三维显示，用户体验更佳。
图3是本发明图像透视畸变校正的方法流程图，应用于包括两个摄像头的拍摄终端，所述方法可通过如图1或图2所示的像透视畸变校正的***100实现，所述方法包括有：
步骤S301，通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像。
所述被摄物体可以是人物、动物、植物、建筑物、山、水、天空等任意一种或多种。所述第一摄像头和第二摄像头为拍摄终端在同一面的两个摄像头，可以是左右摄像头或上下摄像头，所述两个摄像头可同时拍摄且拍摄的内容基本一致，即第一图像和第二图像的图像内容基本一致，存在较多的重合区域。
步骤S302，分别生成第一图像对应的第一三维矢量信息和第二图像对应的第二三维矢量信息。
这里，根据第一图像可以生成第一三维矢量信息，该第一三维矢量信息可以构成第一三维图像；根据第二图像可以生成第二三维矢量信息，该第二三维矢量信息可以构成第二三维图像。本步骤优选的是，生成第一图像中各像素点的第一三维矢量信息，以及生成第二图像中各像素点的第二三维矢量信息。基于像素点的三维矢量信息校正的精度是现有技术通过几张照片获取的几个深度值进行局部大范围校正的精度的千万倍，可大大提高透视畸变校正的速度与精度。即本发明中的三维矢量信息是指图像的平 面二维坐标以及图像中各像素点对应的原像点与摄像头之间距离的矢量信息。所述距离优选为图像的平面二维坐标以及图像中各应像素点对应的原像点与摄像头平面之间距离，当然所述距离也可以是图像的平面二维坐标以及图像中各应像素点对应的原像点与摄像头的中心点之间距离。所述原像点是图像中像素点对应的实际物体的点。
步骤S303，根据得到的第一三维矢量信息和第二三维矢量信息，通过预定的透视畸变校正算法，对第一图像和第二图像进行透视畸变校正。
拍摄终端中可预存有至少一种透视畸变校正算法，例如基于物体距离与方向进行镜头透视畸变校正算法，由于所述透视畸变校正算法为现有技术，因此在此不再赘述。具体而言，根据透视畸变校正算法和第一三维矢量信息，对第一图像进行透视畸变校正；以及根据透视畸变校正算法和第二三维矢量信息，对第二图像进行透视畸变校正。
图4是本发明第一实施例中图像透视畸变校正的方法流程图，应用于包括两个摄像头的拍摄终端，所述方法可通过如图2所示的像透视畸变校正的***100实现，所述方法包括有：
步骤S401，通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像。
步骤S402，分析出第一图像和第二图像的重合区域。所述重合区域的分析涉及到图像识别技术，基于RGB值，亮度，灰度等像素点特征找到相似点与不相似点分界线，即可分析出第一图像和第二图像的重合区域。
步骤S403，计算第一图像在重合区域中每个像素点对应的原像点距第一摄像头的距离，得到第一距离，并根据每个像素点的第一二维坐标和第一距离，生成第一图像的第一三维矢量信息。
即本发明中的第一三维矢量信息是指第一图像的平面二维坐标以及第一图像中对应像素点与第一摄像头之间距离的矢量信息，第一三维矢量信息可以用(x，y，z)来表示。
步骤S404，根据第一三维矢量信息调用预定的第一透视畸变校正参数，并通过透视畸变校正算法对第一图像的重合区域进行透视畸变校正。
步骤S405，计算第二图像在重合区域中每个像素点对应的原像点距第 二摄像头的距离，得到第二距离，并根据每个像素点的第二二维坐标和第二距离，生成第二图像的第二三维矢量信息。
即本发明中的第二三维矢量信息是指第二图像的平面二维坐标以及第二图像中各像素点对应的原像点与第二摄像头之间距离的矢量信息，第二三维矢量信息可以用(x，y，z)来表示。
步骤S406，根据第二三维矢量信息调用预定的第二透视畸变校正参数，并通过透视畸变校正算法对第二图像的重合区域进行透视畸变校正。
第二透视畸变校正参数与第一透视畸变校正参数可以相同或不同。
步骤S407，接收到图像展示指令。
所述图像展示指令的类型可以是二维图像展示指令或者三维图像展示指令。
步骤S408，判断图像展示指令的类型，若图像展示指令为二维图像展示指令则执行步骤S409，若图像展示指令为三维图像展示指令则执行步骤S410。
步骤S409，若图像展示指令为二维图像展示指令，则截取出校正后的第一图像或第二图像中的重合区域进行展示。
步骤S410，若图像展示指令为三维图像展示指令，则对校正后的第一图像和第二图像进行三维调制和展示。
这里，校正后的第一图像和第二图像同时展示可构成三维图像。
图5是本发明第二实施例中图像透视畸变校正方法的实现原理图，应用于包括左右摄像头的拍摄终端，主要包括：
一、通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像。两个
1、开启两个摄拍照模式，通过显示屏对被摄体进行拍摄构图；
2、具体而言，左右摄像头同时对拍摄体进行拍照，获得左右两幅图像并保存，即图像L(即第一图像)、图像R(即第二图像)；
二、分别生成第一图像对应的第一三维矢量信息和第二图像对应的第二三维矢量信息。
分析出图像L和图像R的重合区域，对于两图像L、R的重合区域(图 5中的雪花点区域)，基于每个像素点调用拍摄终端中预存算法进行几何数学计算，得到图像中每个像素点对应的原像点距摄像头距离z，生成图像的三维矢量信息文档如下：
(x1，y1，z1)；(x2，y1，z2)………………………(xn，y1，zn)
(x1，y2，zn+1)；(x2，y2，zn+2)…………………(xn，y2，z2n)
(x1，ym，z(m-1)n+1)；(x2，ym，z(m-1)n+2)………(xn，ym，zmn)
暂存此文档。这里存在有三个坐标系：图像L的第一坐标，图像R的第二坐标，两个图像重合部分构成的重合图像的第三坐标。其中图像L的第一坐标、图像R的第二坐标是二维坐标系，重合图像的第三坐标是三维坐标系。第三坐标(x，y，z)中的X轴原点为图像L的第一坐标中第一列重合像素点的横坐标x1，而第三坐标(x，y，z)中的Y轴原点为重合图像中最下一行的像素点的纵坐标ym。在每个坐标系中，坐标原点是指每个图像(图像L、图像R或重合图像)的左下角最后一个像素点。记住左右两幅图像L和R的重合部份与不重合部分的临界点如图的横轴坐标x1、x2数据。因为图像R、L的坐标不统一，需要x1或x2来转换，如果以图像L的坐标为基准将用到x1，如果以图像R的坐标为基准将用到x2,本实施例是以图像L的坐标为基准为例说明的。第三坐标的坐标原点是基于第一坐标或第二坐标的坐标原点进行变换而来，转换用到x1或者x2。
三、根据得到的第一三维矢量信息和第二三维矢量信息，通过预定的透视畸变校正算法，对第一图像和第二图像进行透视畸变校正。
根据生成的三维矢量信息调用镜头原厂家提供的透视畸变修正系数表通过校准算法分别对两幅图像中重合区域(图中斜线部分)进行透视畸变校正(非重合区域校正值默认为0)。校正过程中三维矢量信息(x，y，z)相关的畸变系数M对应左摄像头拍摄图像L的像素点，M对应右摄像头拍摄图像R的像素点；最后只保存校正后的左右两幅图像LJ、RJ，删除其他临时数据或文件。
后期若只需二维显示图像，剪切两幅图像中任一幅中的重合区域(图中斜线部分)显示即可；若需进行三维显示，分别用现有三维显示技术对左右图像LJ、RJ进行三维调制与同时播放即可。
综上所述，本发明拍摄终端在拍摄时，通过两个摄像头一次性获取被摄物体的两个图像，并分别生成两个图像的三维矢量信息，可大大缩短现有技术需不同距离分别多次对焦拍摄多个图像获取深度信息的时间；然后调用透视畸变校正算法对三维图像进行透视畸变校正。借此，本发明能够实现对三维图像进行透视畸变校正，并且可大大提高透视畸变校正的速度。优选的是，本发明可通过两个摄像头捕获被摄图像中各像素点的三维矢量信息，并根据各像素点的三维矢量信息进行三维图像的透视畸变校正，相比于现有技术通过几张照片获取的几个深度值进行局部大范围校正的精度，本发明大大提高了透视畸变校正的精度。
当然，本发明还可有其它多种实施例，在不背离本发明精神及其实质的情况下，熟悉本领域的技术人员当可根据本发明作出各种相应的改变和变形，但这些相应的改变和变形都应属于本发明所附的权利要求的保护范围。

Claims (10)

  1. 一种图像透视畸变校正的方法，其特征在于，应用于包括两个摄像头的拍摄终端，所述方法包括有：
    通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像；
    分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息；
    根据得到的所述第一三维矢量信息和所述第二三维矢量信息通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正。
  2. 根据权利要求1所述的方法，其特征在于，所述分别生成第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息的步骤包括：
    生成所述第一图像中各像素点的所述第一三维矢量信息；
    生成所述第二图像中各像素点的所述第二三维矢量信息。
  3. 根据权利要求2所述的方法，其特征在于，所述分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息的步骤包括：
    分析出所述第一图像和所述第二图像的重合区域；
    计算所述第一图像在所述重合区域中每个所述像素点对应的原像点距所述第一摄像头的距离，得到第一距离，并根据每个所述像素点的第一二维坐标和所述第一距离，生成所述第一图像的所述第一三维矢量信息；
    计算所述第二图像在所述重合区域中每个所述像素点对应的原像点距所述第二摄像头的距离，得到第二距离，并根据每个所述像素点的第二二维坐标和所述第二距离，生成所述第二图像的所述第二三维矢量信息。
  4. 根据权利要求3所述的方法，其特征在于，所述根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正的步骤包括：
    根据所述第一三维矢量信息调用预定的第一透视畸变校正参数，并通过所述透视畸变校正算法对所述第一图像的所述重合区域进行透视畸变校正；
    根据所述第二三维矢量信息调用预定的第二透视畸变校正参数，并通过所述透视畸变校正算法对所述第二图像的所述重合区域进行透视畸变校正。
  5. 根据权利要求1～4任一项所述的方法，其特征在于，所述根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正的步骤之后包括：
    若接收到图像展示指令，判断所述图像展示指令的类型；
    若所述图像展示指令为二维图像展示指令，则截取出校正后的所述第一图像或所述第二图像中的所述重合区域进行展示；
    若所述图像展示指令为三维图像展示指令，则对校正后的所述第一图像和所述第二图像进行三维调制和展示。
  6. 一种图像透视畸变校正的***，其特征在于，应用于包括两个摄像头的拍摄终端，所述***包括有：
    图像获取模块，用于通过第一摄像头和第二摄像头同时对被摄物体进行拍摄，获取对应的第一图像和第二图像；
    信息生成模块，用于分别生成所述第一图像对应的第一三维矢量信息和所述第二图像对应的第二三维矢量信息；
    图像校正模块，用于根据得到的所述第一三维矢量信息和所述第二三维矢量信息，通过预定的透视畸变校正算法，对所述第一图像和所述第二图像进行透视畸变校正。
  7. 根据权利要求6所述的***，其特征在于，所述信息生成模块包括：
    第一生成子模块，用于生成所述第一图像中各像素点的所述第一三维矢量信息；
    第二生成子模块，用于生成所述第二图像中各像素点的所述第二三维矢量信息。
  8. 根据权利要求7所述的***，其特征在于，所述信息生成模块包括：
    区域分析子模块，用于分析出所述第一图像和所述第二图像的重合区域；
    所述第一生成子模块，用于计算所述第一图像在所述重合区域中每个所述像素点对应的原像点距所述第一摄像头的距离，得到第一距离，并根据每个所述像素点的第一二维坐标和所述第一距离，生成所述第一图像的所述第一三维 矢量信息；
    所述第二生成子模块，用于计算所述第二图像在所述重合区域中每个所述像素点对应的原像点距所述第二摄像头的距离，得到第二距离，并根据每个所述像素点的第二二维坐标和所述第二距离，生成所述第二图像的所述第二三维矢量信息。
  9. 根据权利要求8所述的***，其特征在于，所述图像校正模块包括：
    第一校正子模块，用于根据所述第一三维矢量信息调用预定的第一透视畸变校正参数，并通过所述透视畸变校正算法对所述第一图像的所述重合区域进行透视畸变校正；
    第二校正子模块，用于根据所述第二三维矢量信息调用预定的第二透视畸变校正参数，并通过所述透视畸变校正算法对所述第二图像的所述重合区域进行透视畸变校正。
  10. 根据权利要求6～9任一项所述的***，其特征在于，还包括：
    指令接收模块，用于对所述第一图像和所述第二图像进行透视畸变校正之后，若接收到图像展示指令，判断所述图像展示指令的类型；
    第一展示模块，用于若所述图像展示指令为二维图像展示指令时，截取出校正后的所述第一图像或所述第二图像中的所述重合区域进行展示；
    第二展示模块，用于若所述图像展示指令为三维图像展示指令时，对校正后的所述第一图像和所述第二图像进行三维调制和展示。
PCT/CN2015/080314 2015-03-31 2015-05-29 图像透视畸变校正的方法及*** WO2016155110A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510149800.4A CN104867113B (zh) 2015-03-31 2015-03-31 图像透视畸变校正的方法及***
CN201510149800.4 2015-03-31

Publications (1)

Publication Number Publication Date
WO2016155110A1 true WO2016155110A1 (zh) 2016-10-06

Family

ID=53912930

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/080314 WO2016155110A1 (zh) 2015-03-31 2015-05-29 图像透视畸变校正的方法及***

Country Status (2)

Country Link
CN (1) CN104867113B (zh)
WO (1) WO2016155110A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117115144A (zh) * 2023-10-18 2023-11-24 深圳市强达电路股份有限公司 一种pcb内部孔位缺陷在线检测***

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105704374B (zh) * 2016-01-29 2019-04-05 努比亚技术有限公司 一种图像转换装置、方法和终端
CN107144241B (zh) * 2017-06-09 2019-01-01 大连理工大学 一种基于景深补偿的双目视觉高精度测量方法
CN109146977A (zh) * 2017-06-16 2019-01-04 中兴通讯股份有限公司 图像平移误差修正方法、移动终端及计算机可读存储介质
CN107748887A (zh) * 2017-09-30 2018-03-02 五邑大学 一种基于显性与隐性线段检测的平面文档图像透视矫正方法
CN107680060A (zh) * 2017-09-30 2018-02-09 努比亚技术有限公司 一种图像畸变校正方法、终端及计算机可读存储介质
CN108171744A (zh) * 2017-12-26 2018-06-15 努比亚技术有限公司 一种双目虚化中视差图的确定方法、移动终端及存储介质
CN108197624A (zh) * 2018-02-02 2018-06-22 杭州清本科技有限公司 证书图像校正识别方法及装置、计算机存储介质
CN113826376B (zh) * 2019-05-24 2023-08-15 Oppo广东移动通信有限公司 用户设备和斜视校正方法
CN110072045B (zh) * 2019-05-30 2021-11-09 Oppo广东移动通信有限公司 镜头、摄像头及电子设备

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120194685A1 (en) * 2011-01-31 2012-08-02 Canon Kabushiki Kaisha Imaging device detecting motion vector
CN103065289A (zh) * 2013-01-22 2013-04-24 清华大学 基于双目立体视觉的四目摄像机正面人脸重建方法
CN103077521A (zh) * 2013-01-08 2013-05-01 天津大学 一种用于视频监控的感兴趣区域提取方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004040395A (ja) * 2002-07-02 2004-02-05 Fujitsu Ltd 画像歪み補正装置、方法及びプログラム
CN102005039B (zh) * 2010-08-24 2012-05-23 浙江大学 基于泰勒级数模型的鱼眼相机立体视觉深度测量方法
CN103824303A (zh) * 2014-03-14 2014-05-28 格科微电子（上海）有限公司 基于被摄物的位置、方向调整图像透视畸变的方法和装置
CN104259669A (zh) * 2014-09-11 2015-01-07 苏州菲镭泰克激光技术有限公司 一种精准的三维曲面激光标刻方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120194685A1 (en) * 2011-01-31 2012-08-02 Canon Kabushiki Kaisha Imaging device detecting motion vector
CN103077521A (zh) * 2013-01-08 2013-05-01 天津大学 一种用于视频监控的感兴趣区域提取方法
CN103065289A (zh) * 2013-01-22 2013-04-24 清华大学 基于双目立体视觉的四目摄像机正面人脸重建方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117115144A (zh) * 2023-10-18 2023-11-24 深圳市强达电路股份有限公司 一种pcb内部孔位缺陷在线检测***
CN117115144B (zh) * 2023-10-18 2024-05-24 深圳市强达电路股份有限公司 一种pcb内部孔位缺陷在线检测***

Also Published As

Publication number Publication date
CN104867113B (zh) 2017-11-17
CN104867113A (zh) 2015-08-26

Similar Documents

Publication Publication Date Title
WO2016155110A1 (zh) 图像透视畸变校正的方法及***
EP3163535B1 (en) Wide-area image acquisition method and device
CN110717942B (zh) 图像处理方法和装置、电子设备、计算机可读存储介质
WO2018068719A1 (zh) 一种图像拼接方法及装置
CN111862224B (zh) 确定相机与激光雷达之间外参的方法和装置
US20170214866A1 (en) Image Generating Method and Dual-Lens Device
CN109474780B (zh) 一种用于图像处理的方法和装置
US9946955B2 (en) Image registration method
WO2018153313A1 (zh) 立体视觉摄像机及其高度获取方法、高度获取***
US11282232B2 (en) Camera calibration using depth data
CN110689581A (zh) 结构光模组标定方法、电子设备、计算机可读存储介质
CN108629756B (zh) 一种Kinectv2深度图像无效点修复方法
CN108470356B (zh) 一种基于双目视觉的目标对象快速测距方法
CN109656033B (zh) 一种区分液晶显示屏灰尘和缺陷的方法及装置
CN109920004B (zh) 图像处理方法、装置、标定物组合、终端设备及标定***
KR101589167B1 (ko) 깊이 정보를 이용한 원근 왜곡 영상의 보정 시스템 및 방법
WO2018001252A1 (zh) 一种投射单元及包括该单元的拍摄装置、处理器、成像设备
CN112470192A (zh) 双摄像头标定方法、电子设备、计算机可读存储介质
US11523056B2 (en) Panoramic photographing method and device, camera and mobile terminal
TWI615808B (zh) 全景即時影像處理方法
CN103778607A (zh) 一种图像校正方法
US20240022702A1 (en) Foldable electronic device for multi-view image capture
TWI569642B (zh) 機器視覺中進行影像擷取之方法及裝置
WO2020146965A1 (zh) 图像重新聚焦的控制方法及***
CN116912331A (zh) 标定数据生成方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15887077

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15887077

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 17.04.2018)

122 Ep: pct application non-entry in european phase

Ref document number: 15887077

Country of ref document: EP

Kind code of ref document: A1