78 results about "Vanishing line" patented technology

A method reconstructs at three-dimensional (3D) real-world scene from a single two-dimensional (2D) image by identifying junctions satisfying geometric constraint of the scene based on intersecting lines, vanishing points, and vanishing lines that are orthogonal to each other. Possible layouts of the scene are generated by sampling the 2D image according to the junctions. Then, an energy function is maximized to select an optimal layout from the possible layouts. The energy function use's a conditional random field (CRF) model to evaluate the possible layouts.

A vehicle-mounted detecting device and corresponding method for detecting a distance between the vehicle and an object outside the vehicle. The detecting device extracts edge intersection points from images of individual frames photographed by a camera. A set of edge intersection points for which the relative position in images does not change between frames is then specified as an observation point group; the movement of the specified observation point group in images is observed; and pitching by the vehicle is detected. Any two points with differing heights on an object present in an image are extracted as feature points, and after the effect of pitching is eliminated from the feature points, the position of a vanishing line in the image is calculated based on the ratio of the velocities of the feature points in images after the length between the two feature points and the effects of pitching are eliminated.

A method of dynamically ascertaining the alignment of a vehicular camera. The method involves acquiring a sequence of images provided by the camera whilst the vehicle is in motion. For each range of steering angles, the method (i) selects a plurality of feature points in the images, (ii) tracks a motion trajectory for each selected feature point, and (iii) determines a vanishing point in the image plane based on the tracked motion trajectories. The method determines a vanishing line in the image plane based on a locus of these vanishing points and determines the alignment of the camera based on the position of a central vanishing point (corresponding to the zero degree angle) and the vanishing line.

The invention discloses a method for creasing a distance map based on monocular camera machine vision, which comprises the following steps: carrying out red component extraction on an image collected by a car camera for obtaining a baseline image; utilizing an edge detection operator for carrying out boundary detection on the baseline image, and further carrying out Hough transform on discontinuity points in the boundary for detecting a boundary line; carrying out the Hough transform on the boundary detection result for finding out straight lines in the image, then carrying out statistical weighting on end points of a plurality of straight lines approaching a vanishing point, taking the average value of detected column coordinates of a group of the straight lines as the column coordinate of the vanishing point of an optical axis of the camera, and taking the average value of row coordinates as the row coordinate position of a vanishing line; and respectively filling distance information from a projection line of the optical axis to the boundaries on the two sides, carrying out approximate treatment on scenes being far away from the vanishing line, eliminating the impacts of long-distance errors on defogging treatment and obtaining the distance-depth map of the scenes. The utilization of the method can solve the problems of difficult creation of the distance map and expensive distance measuring equipment during model defogging.

A system and method of generating three-dimensional (3D) depth information is disclosed. The vanishing point of a two-dimensional (2D) input image is detected based on vanishing lines. The 2D image is classified and segmented into structures based on detected edges. The classified structures are then respectively assigned depth information.

The invention provides a vision-based vehicle detection, tracking and early warning method. The method comprises the steps of collecting the image and calibrating the road vanishing line; defining andgraying the vehicle detection area, and then performing gray stretching on the vehicle detection area image according to the illumination intensity classification of the collected image; constructingthe training sample images and labeling the images as positive sample images and negative sample images manually; extracting the haar feature and LBP feature of the positive sample images and negative sample images to train an Adaboost cascade classifier; dividing the vehicle detection area into different domains, and detecting the vehicle by the Adaboost cascade classifier after training, and judging the vehicle twice according to the illumination intensity. When the vehicle is detected, the KCF target tracking method is used to track the vehicle. The vehicle is tracked, and the distance between the vehicle and the vehicle is calculated by the distance estimation method based on the position, and the collision time is calculated according to the vehicle speed and the distance between thevehicle and the vehicle to give the early warning. The method reduces the computational complexity and improves the vehicle detection accuracy.

The invention discloses a height measurement method based on video multi-target tracking. The method includes the following steps that background modeling is conducted on video sequences collected by a camera, and foreground images are extracted through background subtraction; each frame image is mapped into an undirected network diagram G=<V, E>; an energy function is built; the built energy function is minimized, and label values of pixels of the current frame image are acquired, wherein the pixels belong to different targets and backgrounds,; different colors are given to the pixels which belong to the different targets, and a tracking frame of the multiple targets is determined; a vanishing point Vy where the camera is perpendicular to the horizontal plane and a vanishing line l of the horizontal plane are calculated; head feature points and foot feature points of the target to be detected in each frame image are extracted; the height of the targets to be detected in each frame image is calculated; the height measurement results of the multi-frame video sequences are merged, and the actual height of the targets to be detected is determined. According to the method, the camera does not need to be completely marked, only the vanishing point and the vanishing line of the horizontal plane need to be calculated, and therefore calculation complexity is reduced.

A vehicle-mounted detecting device and corresponding method for detecting a distance between the vehicle and an object outside the vehicle. The detecting device extracts edge intersection points from images of individual frames photographed by a camera. A set of edge intersection points for which the relative position in images does not change between frames is then specified as an observation point group; the movement of the specified observation point group in images is observed; and pitching by the vehicle is detected. Any two points with differing heights on an object present in an image are extracted as feature points, and after the effect of pitching is eliminated from the feature points, the position of a vanishing line in the image is calculated based on the ratio of the velocities of the feature points in images after the length between the two feature points and the effects of pitching are eliminated.

The present invention relates to a method for calibrating a paracatadioptric camera using the images of a single sphere and circular points. According to the method, two different points are selected from the image of the sphere, so that two vanishing points are obtained; a straight line where the two vanishing points are located is a vanishing line, the images of the vanishing line and the image of the sphere intersect with the images of the circular points, and three images can provide the images of three circular points; and the inner parameters of the camera are solved through using the constraint of the images of the circular points for an absolute conic curve image. The method specifically includes the following steps of: fitting a mirror surface contour projection equation and a target projection equation; estimating an antipodal sphere image; determining the images of the circular points; and solving the inner parameters of the paracatadioptric camera. In a process of solving the inner parameters of the paracatadioptric camera, five inner parameters of the paracatadioptric camera are solved by using three images of a target shoot by the paracatadioptric camera.

A low-complexity method of converting 2D images / videos into 3D ones includes the steps of identifying whether each pixel in one of the frames is an edge feature point; locating at least two vanishing lines in the frame according to the edge feature point; categorizing the frame into the one of close-up photographic feature, of landscape feature, and of vanishing-area feature; if the frame is identified to have the vanishing-area feature or the landscape feature to generate a GDM; and apply a modificatory procedure to the GDM to generate a final depth information map; if the frame is identified to have the close-up photographic feature, distinguish between a foreground object and a background information in the frame and define the depth of field to generate the final depth information map.

The invention discloses a lane line detection method, and belongs to the technical field of image processing. The lane line detection method comprises the steps that first a pre-processed image is horizontally divided into K edge image blocks, the height scale of the edge image block at the bottom end and the entire image is [1 / 4,1 / 3], and two or more pairs of straight lines which are closest to a vanishing line are determined under the determined vanishing line to be used as candidate line pairs of the current edge image block; then the weights of the candidate line pairs are calculated, the straight line with the maximum weight of the edge image blocks of the candidate line pairs is taken as the only lane line segment pair of the current edge image block based on the weights of the candidate line pairs; at last, the lane line of a current frame image is output based on the end points of the lane line segment pairs of all the edge image blocks. The method is used for detecting lane lines, the method is not sensitive to initialization parameters, the robustness of detection is high, and a good detecting effect can be achieved under the bad conditions of dark light, lane line loss, shadows and the like.

The invention relates to a pose measuring method based on coaxial circle characteristics of a target and belongs to the technical field of computer vision measurement. The pose measuring method is characterized in that the target is provided with coaxial circle characteristics, curve extraction and ellipse fitting technologies are combined to obtain two coaxial circular projection equations in an image, and circular point projection coordinates and vanishing line equation of a target plane are obtained by means of relation of absolute conic, vanishing line and circular points in projection geometry of the circle characteristics; by means of the polar line-pole theorem, circular center projection coordinates of two coaxial circles are obtained by the vanishing line equation of the plane, and pose and position information of the target is solved by taking the actual distance of the coaxial circles as prior condition and combining the projection of the circular points and the projection coordinates of the circular centers. The pose measuring method based on the coaxial circle characteristics of the target has the advantages that measuring of target pose is completed by means of a single target image of coaxial circle characteristics, so that the pose measuring method is simple to operate and applicable to measurement in real time. Meanwhile, without manual intervention, high-precision measurement is realized.

The present invention relates to a method for calibration of a parabolic catadioptric camera by employing a ball image and common self-polar triangles. A ball in a space is employed as a calibration object to fit a mirror surface contour projection equation and a ball image equation under a parabolic catadioptric camera and calculate an antipodal ball image equation and an equation of an image ofa great circle having a center of a circle being coincided with a center of a circle of a unit virtual ball and being parallel with a projection small circle of the ball on a unit visual ball model, two common poles, corresponding to two common self-polar triangles formed by the ball image and the antipodal ball image, of the image of the great circle are connected with two common poles to solve avanishing line, a relation of the projection of the circle and the vanishing line is employed to solve the images of circular ring points, and three images are shot to obtain images of the three circular ring points, and the images of the circular ring points and the constraint of the internal parameter matrix of the camera are employed to complete calibration of the camera.

The invention relates to a method for solving camera inner parameters by utilizing a bimirror device and circular point characteristics. The bimirror device is specifically composed of two rectangular plane mirrors, the two plane mirrors are perpendicular to a plane formed by bottom edges of the two plane mirrors, and the bimirror device is used for a camera self-calibration device. The method includes the steps of shooting three images including an image of an object and four images formed in the plane mirrors of the object, extracting characteristic points from the images, arranging one set of corresponding characteristic points on one circle, enabling points at infinity in the direction of the two real plane mirrors to be on the same straight line, calling the straight line as an infinity straight line, calling images of the infinity straight line as a vanishing line, regarding the intersection point of the projection of the circle and the vanishing line as the image of a circular point, setting six sets of constraint equations of the image of the circular point on an absolute conic image by utilizing the three images, and solving the camera inner parameters in a linear mode.

The invention provides an automatic calibration method of a camera and an intelligent video monitoring method based on automatic calibration of the camera. The intelligent video monitoring method comprises the following two steps of: firstly, automatically calibrating the camera; and applying a calibration result in a video monitoring image, and calculating all relative target information to realize intelligent video monitoring. The intelligent video monitoring method disclosed by the invention can automatically calibrate by utilizing the camera, and avoids operation of manually marking reference points, a package box can be replaced by a minimum compact package box, thus being capable of improving the precision of a calculation result, reducing manual mark in a calibration process relative to a geometric model based on a vanishing point and a vanishing line, and being simple in operation.

The invention relates to a method and a device for identifying road direction, and the method comprises the following steps: calculating the direction of each candidate lane line in reality extracted from the part below a vanishing line in an image absorbed by a vehicle-borne image absorption device; calculating the total length of the candidate lane line of area D in each direction of multiple directions in the image, while the multiple direction areas are obtained by dividing the directions in the image and the total length of the candidate lane line of the directional area D is equal to the length sum of the candidate lane line of each candidate lane line in reality located in the directional area D; determining the directional area with the calculated longest total length of the candidate lane line in the multidirectional areas as the road direction of the image. By using the method and the device in the invention, road direction can be detected based on the candidate lane line without accurately identifying the lane line.

The invention discloses a global calibration method of a distributed multi-camera system. The global calibration method includes the steps of using two-dimensional calibration targets formed by two groups of mutually orthogonal parallel straight lines to obtain relative position and attitude initial values through vanishing line equations and known target geometric dimensions based on vanishing point and vanishing line estimation cameras and relative positions and attitudes of the targets, and seeking minimum values of re-projection error functions so as to obtain optimum values of the cameras and the calibration targets; using auxiliary cameras to shoot adjacent target images to obtain a coordinate transformation matrix between the adjacent targets, seeking relative position and attitude initial values of the calibration targets relative to a reference target by an increment method and obtaining optimum values of transformation matrixes of the calibration targets relative to the reference target by seeking minimum values of re-projection error functions based on closed image sequences; through corresponding coordinate transformation, seeking optimum values of the transformation matrixes of the cameras relative to a reference camera to conduct global calibration of the distributed multi-camera system.

The invention relates to a method for calibrating a parabolic refraction and reflection camera through the Veronese mapping and a checkerboard. The method is characterized in that the target of the method is formed by a known checkerboard in the space. The method comprises the specific steps of during the solving process of the internal parameters of a parabolic refraction and reflection camera, shooting 3 images by the parabolic refraction and reflection camera; obtaining three homography matrixes through the Veronese mapping so as to solve a vanishing point for any straight line; solving another vanishing point based on the relationship between antipodal points and harmonic conjugates, wherein the connecting line of the two vanishing points is a vanishing line; according to the intersection points of the vanishing line with the image of the straight line namely a quadratic curve, solving the image of three sets of circular points; based on the constraint of the image of circular points on the image of the absolute quadratic curve, solving the internal parameters of the parabolic refraction and reflection camera through the Cholesky decomposition.

The invention discloses a parameter estimation method, a parameter estimation apparatus and electronic equipment. The parameter estimation method includes the following steps: acquiring a current frame of image and a pre-frame of image acquired by an imaging device; determining the mapping relation between the current frame of image and the pre-frame of image; and according to the mapping relation and the pre position coordinates of the vanishing line in the pre-frame of image, estimating the current position coordinates of the vanishing line in the current frame of image. Therefore, the parameter estimation method and the parameter estimation apparatus can estimate the position coordinates of the vanishing line in the current frame of image acquired by the imaging device, with higher precision, and can satisfy the vanishing line estimation requirement in real time.

The invention discloses a method for calibrating a parabolic catadioptric camera by using properties of a single sphere and a circle tangent line. The method is characterized by taking a sphere in a space as a target. The method comprises the specific steps of: firstly, shooting three pictures containing the sphere from different positions by means of the parabolic catadioptric camera; secondly, acquiring antipodal image points according to relationship between image points and antipodal image points thereof in a parabolic catadioptric unit visual sphere model, so as to estimate an antipodal sphere image of a sphere image; thirdly, selecting a point from the sphere images, wherein a tangent line of each image point about the corresponding sphere image and a tangent line of the corresponding antipodal image point about the corresponding antipodal sphere image intersect at an image vanishing point through definition of the antipodal image points, an intersection point of the image vanishing point about polar lines of the sphere image and the antipodal sphere image is an image vanishing point, a straight line where the two image vanishing points are located is an image vanishing line,an intersection point of the image vanishing line and the sphere image is an image of circular points, and the three pictures provide the images of three groups of circular points; and finally, solving inner parameters of the camera by utilizing the constraint of the images of the circular points for an absolute conic curve image.

Techniques pertaining to methods and devices for calibrating camera parameters are disclosed. According to one aspect of the present invention, objects on an image are detected and features on the objects, such as the top of a person, are identified. Perspective parallel lines based on the object features are constructed and vanishing points from the converged perspective parallel lines are determined. A vanishing line is defined according to the vanishing points. The intrinsic parameters, such as focal length, and the extrinsic parameters, such as tilting and pan angles, of a camera are calibrated according to the vanishing points, the vanishing line and a reference object with known dimensions, or known camera angles, The calibrated camera parameters are used for accurate photogrammetric measurements and computer vision applications.

The invention discloses a highway and urban road lane line identification method. The highway and urban road lane line identification method comprises the following steps of: obtaining a lane image through a camera, obtaining internal and external parameters of the camera by utilizing a camera parameter self-calibration method, and obtaining a lane plane vanishing line; performing gray-scale treatment of the lane image, and dividing an interested area; performing median filtering of the image obtained in the previous step; for the median filtered image, extracting textural features of a lane line by using Gabor transformation; describing edge features of the lane line by using multi-angle Haar features, and performing classified identification by utilizing an Adaboost classifier; for a structured road, designing a hyperbola combination model; and estimating parameters of the hyperbola combination model by utilizing an improved Ransac algorithm. By means of the highway and urban road lane line identification method disclosed by the invention, the lane line detection problem in complex environments, such as shadow, abrasion of the lane line and severe weather, can be solved; and furthermore, the highway and urban road lane line identification method has better real-time performance.

The invention relates to a method for solving intrinsic parameters of a camera by circular motion in a double-plane-mirror device. The device is composed of two plane mirrors and a pinhole camera, and space points and reflection image points thereof in the plane mirrors are located on a same space circle. The method includes: shooting an object and three images of four reflection images of the object in the plane mirrors, extracting features points from the images, and fitting a curvilinear equation; acquiring an image of the circle center by the aid of symmetrical property of reflection point groups relative to the plane mirrors and tangential property of a curve at the feature points, and then utilizing relation among poles and polar lines to obtain a vanishing line on an image plane, namely an image of a line at infinity; acquiring an interaction of the vanishing line and the curve, namely an image of a circular point, creating six constraint equations of the image of the circular point relative to an absolute conic image, and linearly solving intrinsic parameters of the camera.

A method of obtaining a depth map from a single digital image of a three-dimensional scene containing geometric objects is provided that includes the identification in the image of at least one horizontal plane or one vertical plane of the three-dimensional scene and the assignment to every point of the image forming part of the identified plane, or to each one of the identified planes, of a depth level according to its distance from, respectively, a predetermined horizontal or vertical reference straight line of the image. In particular, the identification of at least one plane includes the identification in the image of at least two vanishing lines corresponding to two parallel lines in the plane and the assignment of a depth level made in accordance with the angular coefficients of the identified vanishing lines. Also included is a completely automatic method for obtaining a depth map from a single digital image of any three-dimensional scene.

The present invention discloses a method for calibrating a parabolic reflex camera by using the properties of conjugate diameters of a straight line and a circle. Firstly, an edge point of a target image and an edge point of a mirror contour projection of one image are extracted from three images, and the mirror contour projection is obtained by least square fitting. A point is taken on a line image, and an antipodal image point is obtained. An intersection of a tangent of the image point to a line image and a tangent of the antipodal image point to the line image is a vanishing point. Two heterogeneous points are taken on the line image, two vanishing points are obtained, and a vanishing line is determined by the two vanishing points. An intersection of a polar of the vanishing points tothe line image and the vanishing line and the vanishing points form a group of orthogonal vanishing points, and six groups of orthogonal vanishing points are provided by the three images. Finally, theconstraint of an absolute quadratic curve image with the orthogonal vanishing points is used to solve camera internal parameters.

The invention discloses a camera self-calibration method based on an absolute conic image, belonging to the technical field of image processing and camera calibration. The method comprises the following steps: extracting feature points and feature lines, calculating images of vanishing points, vanishing lines and imaginary circles; according to the image of the vanishing point and the imaginary circle, the image of the absolute quadratic curve is calculated and the intrinsic parameter matrix of the camera is obtained. The rotation matrix is calculated by using the orthogonal property of the rotation matrix, and the translation vector is calculated by using the position of the camera center in the world coordinate system. The Euler angle of the camera is calculated by comparing the erasingline calculated by perspective projection model with the erasing line obtained by fitting the erasing points. Normalization algorithm is used to solve the basic matrix, determine the constraint relationship between different images, and complete camera self-calibration. The invention has the advantages of designing a new target, which is simple, convenient and practical to calibrate, has wide practicability, improves the accuracy of data processing, and reduces the influence of coordinate transformation on data.

The invention is suitable for the technical field of image processing, and provides a camera calibration method and device, electronic equipment and a storage medium, and the camera calibration methodcomprises the steps: obtaining a first image, photographed by a camera, of a calibration plate at a first preset position and a second image, photographed by the camera, of the calibration plate at asecond preset position, acquiring a second preset position by rotating the calibration plate around a preset rotating shaft for a preset angle at the first preset position, wherein the preset rotating shaft is perpendicular to a preset plane, detecting an angular point in the first image and an angular point in the second image, and calculating a homography matrix according to the angular point in the first image and the angular point in a second image, using the homography matrix for describing the mapping relation between the second image and the first image, calculating the vanishing lineof the preset plane according to the homography matrix so that the calculation accuracy of the vanishing line of the preset plane is improved, calibrating the camera according to the vanishing line ofthe preset plane so that calibration precision of the camera is improved.

The invention belongs to image processing technology, and relates to an image manipulation detection method based on single view measurement. The method comprises the steps of: finding out a horizontal vanishing line 1 infinity and two vanishing points vx and vy which are vertical to each other for a given image; manually labeling two end points of a reference object and a target object on the image; calculating the height of the target object; and judging whether the target object is a manipulated area by judging whether the height is reliable, or in an accepted range or not. The method can realize the authenticity verification of a digital media (digital image or video) by utilizing little geometric information.

The invention discloses a self-calibration method of multi-view structured light, comprising the steps of: (1) projecting a grid pattern by a projector, carrying out code generation by the grid pattern on the basis of a position only principle, and acquiring three pictures by a video camera on three non-collinear positions, wherein a multi-view structured light system comprise the video cameral and the projector; and (2) defining Kc to be an inner parameter matrix for the video camera, phi 1 and phi 2 to be two arbitrary projector planes, Hi to be a homography matrix obtained by the video camera through transmission via the plane phi 1 from c1 to c2 and from c2 to c3, i to be 1 or 2, and Kct to be a pole of a video camera c1 and to be equal to t3e2; and determining a vanishing line of a stripe plane of the projector in a first video camera coordinate system through the homography matrix obtained through transmission via the strip plane, and obtaining the inner parameter matrixes of the video camera and the projector by combining infinite homography knowledge. The method can simplify calculation, has good applicability and benefits cost reduction.