45 results about "Multi camera calibration" patented technology

The invention discloses a high-precision motion platform-based multi-camera calibration method. In a whole working region, a calibration plate fixed to a high-precision motion platform through an objective table can make any known linear motion of rotation, translation and the like, and a pose of the calibration plate in a world coordinate system is obtained. The calibration plate moves under the control of the high-precision motion platform to appear in sight of one or more cameras, the cameras shoot images, coordinates of corresponding points in the world coordinate system and an image coordinate system are searched for, and finally parameters of all the cameras are solved once by adopting a global optimization algorithm, so that the shortcoming of low efficiency, relatively large accumulative error, trouble of designing a specific calibration part or the like due to one-by-one calibration of cameras in a multi-camera system in an existing calibration method is overcome, one-time calibration of the whole multi-camera system is realized, the method has greatly important engineering application values, and high-precision calibration and three-dimensional reconstruction are finished. A measurement process is simple, a calibration result is accurate, and the working efficiency is improved.

The invention discloses a multi-camera calibration and alignment fitting method. A posture of a target object is observed by using a first camera group; a posture of a fitting object is observed by using a second camera group; and rotating coordinate systems of the two camera groups coincide. A calibration process comprises a first calibration stage, a second calibration stage and a third calibration stage. In the first calibration stage, internal parameters and a camera mounting angle of each camera in the camera groups are calibrated; in the second calibration stage, the position, in an output coordinate system of each camera in the camera groups, of a rotating center of a correction platform is calibrated and the rotating coordinate systems are established according to the condition that the rotating center of the correction platform is a fixed point after returning to zero; and in the third calibration stage, mapping and calibration are performed: the first camera group is mapped and calibrated by the second camera group according to the condition that the rotating coordinate systems of the first camera group and the second camera group coincide.

Passive wands, tracked by one or more cameras, are used as user interface devices. The wands are inexpensive, and are a natural way for users to interact with devices such as large displays. Each wand may be used to specify a full 6 degrees of freedom, which can be helpful for manipulation of 3D applications. The wands may also be used to identify 3D world points, which is useful for multi-camera calibration, and for building 3D models of physical spaces. The described methods provide a more precise estimation of wand pose, and can be implemented with a single camera. Also described is a method for estimation wand pose using information available along a one dimensional ‘line scan’ within the image, which simplifies processing.

The invention discloses a three-dimensional bent-pipe multi-camera visual inspection method and a system thereof. The method comprises a step S1 of multi-camera calibration, a step S2 of bent-pipe image acquisition, a step S3 of bent-pipe edge identification, a step S4 of two-dimensional bent-pipe center line detection, a step S5 of two-dimensional bent-pipe center line reconstruction, and a step S6 of bent-pipe parameter calculation. Due to the operation steps, a quick and accurate measurement method is provided for the dimension detection of a machined bent pipe, and the method is not only convenient to operate and high in automation degree, and also accurate and reliable in measurement results.

The invention puts forward a vehicle video speed measurement method based on a multi-viewpoint camera, and is used for solving the problem of big error of vehicle speed detection since the measurementdistance of a single camera is unstable. The method comprises the following steps that: calibrating a plurality of cameras; carrying out rectification on an image coordinate and a world coordinate; according to multiple viewpoints, positioning a vehicle target; and according to positioning data, measuring the movement distance and the movement speed of the vehicle. By use of the method, through multi-camera calibration, monitoring picture collection and a method based on statistical learning, the target is accurately positioned, vehicle speed is calculated to realize the speed measurement ofthe video, the positioning accuracy of the target vehicle in video collection data is improved, in addition, an inspection missing problem in the single-camera target detection is avoided and reduced,the calculation amount of the image coordinate and the world coordinate is small, the real-time detection of video monitoring can be realized, and an alarm is given. The method is high in measurementand wide in applicable ranges, a situation that single camera detection is not integral can be avoided, and a severe environment is coped.

The invention discloses a method for reconstructing real-time three-dimensional face data with multiple images. The method includes the following steps: constructing shooting systems of eight cameras; performing multi-camera calibration; performing image matching, to be more specific, calculating global matching points of an ith group of images, and calculating accurate matching points of the ith group of images; calculating three-dimensional face coordinates via the matching points and a projection matrix of the cameras, to be more specific, imaging an object point on the third image of the ith group, and taking an intersection of three projection lines as the object point so as to complete reconstructing of the real-time three-dimensional face data. The face data obtained in the method are global, dense and real-time and can be used directly, and the method can be widely applied to the fields such as the animation industry, computer 3D (three-dimensional) games, human-computer interaction, identity recognition, medical virtual surgery, videophones, face recognition, face expression, age simulation, film and TV advertising production and computer cognition.

Techniques of augmented reality device calibration are disclosed. In some example embodiments, a system calibrates a visual inertial navigation (VIN) camera of a head mounted display (HMD) device with at least one eye camera of the HMD device to generate multi-camera calibration parameters, calibrating the at eye camera(s) with a display module of the HMD to generate display calibration parameters, calibrating the IMU with the VIN camera to generate VIN calibration parameters, and calibrating the IMU to the display module using the multi-camera calibration parameters, the display calibration parameters, and the VIN calibration parameters.

The invention relates to a virtual multi-camera target tracking video material generation method. By aiming at situations that research video source materials are in lack and scene changes are insufficient for multi-camera target tracking, the invention utilizes a vision technology to truly simulate a complex environment of the multi-camera target tracking in a real environment by creating a required scene and customizing the attributes of a tracked target and cameras, and even tracking video materials which rarely appear in the reality can be created. The virtual multi-camera target tracking video material generation method favorably simulates problems of specific complex environments, such as shadows, target occlusion, illumination variation, target deformation, texture changes, target generation and disappearing and the like which are not easy to realize in the reality in a target tracking process, and the three-dimensional realization processes of multi-camera calibration, collaborative operation, synchronous calibration, network layout and the like are solved. The frame data of a camera channel is processed to generate a virtual multi-camera target tracking vide, and great help is provided for target tracking algorithm research and verification.

The invention discloses a multi-camera calibration method based on a cylindrical calibration object. The method is mainly composed of chessboard grids and coded patterns, cylinder calibration and a multi-camera calibration. The method has the advantages that (1) using a standard cylindrical object as a calibration object, achieving single-camera calibration and easy operation, saving operation time and achieving high calibration precision; (2) using an annular coded marker to complete acquisition while acquiring a camera calibration image so as to provide convenience for multi-camera calibration while determining the positions and the sequence of the cameras; (3) accurately achieving multi-camera global calibration and saving operation steps and time by using the result of single-camera calibration and the uniquely determined camera positions and sequence.

The invention discloses an acquisition method of large scene based on line laser, including following steps: 1) erecting a plurality of cameras around the scene and regulating focal distance; 2) applying multi-camera calibration algorithm to calibrate and calculate internal and external parameter of a plurality of cameras; 3) using line laser A and B of emitting point A and B from outside the scene to scan the whole scene, all the cameras screening the line laser A and B of different time respetively in the scanning process to obtain a plurailty groups of pictures having information of both the line laser A and B of different time; 5) extracting line laser from the screened pictures; 6) proceeding three-dimensional point cloud calculation to the extracted line laser to generate three-dimensional information of the whole scene. Method of the invention avoids the turntable calibration step by improving the conventional calculation, which both improving scan accuracy and recuding scene acquisition time, and improving scan efficiency. The method is suitable to scan complex scenes incable of fixing laser emitter.

The invention discloses a multi-camera calibration and parameter optimization method based on a sphere projection common tangent line. The method includes obtaining the coordinates of the sphere center projection on the phase plane according to the collinear principle of the sphere projection common tangent line intersection point and the sphere center projection point; 2, through the combinationwith the sphere center projection, obtaining the inner parameters and the outer parameters of the camera through the algebraic relation between the image of the absolute quadratic curve and the projection of the sphere; 3, representing the projection of the sphere by using the oval parameters, dividing the error between the projection curve and the reprojection of the projection curve into two relatively independent parts accordingly, and adopting a nonlinear optimization method for gradually optimizing the deviation of the two parts to obtain a final calibration result of the camera; 4, unifying the calibration results of the multiple cameras to the same world coordinate system through the matching method, and completing the multi-camera calibration. According to the method, the parametergeometric significance is reserved in the solution process, the calibration precision is high, and the simultaneous calibration of the multiple cameras is simply and efficiently completed.

The invention relates to a single-point calibration object-based multi-camera internal and external parameter calibration method and a calibration component. The calibration method comprises the following steps of: acquiring a single calibration point which moves freely and an image point of an L-shaped rigid body which is used for indicating a world coordinate system by using a plurality of infrared cameras which are fixed at different positions in a scene; and uploading the single calibration point and the image point into an upper computer so as to calibrate internal and external parametersof the plurality of cameras by utilizing image point data. According to the method, cameras with common viewpoints can be calibrated in pairs according to a pinhole and distortion camera model and anepipolar geometric constraint relationship between image point pairs. According to the method, a utilized calibration tool is simple to manufacture, and calibration objects do not need to be limitedto move in a common view field of all the cameras, so that the operability is strong; through a multi-camera cascade path determined by utilizing a common view field relationship, more image points can participate in operation, so that the algorithm robustness is better; and through multi-step optimization, calibration parameters can achieve sub pixel-level re-projection errors, so that high-precision demands can be completely satisfied.

The invention provides a 3D automobile wheel positioner multi-camera calibration system and method. The 3D automobile wheel positioner multi-camera calibration system and method are mainly used for conducting production and maintenance field calibration on the 3D automobile wheel positioner multi-camera system. The information of the relative positions of two target boards is obtained through a calibration camera, two measurement cameras each obtain the position information of one target board, and finally the relative position relation of the second measurement camera relative to the first measurement camera is calculated. An extra target board is not needed in the calibration system, a calibration rod does not need to be used for rigid connection of the two target boards, calibration results are high in accuracy, the size and the weight of the calibration system are greatly reduced, maintainers can conveniently carry the system, the calibration steps are simple, operation is convenient, and relevant cost is reduced.

The invention discloses an ORB-SLAM-based multi-camera calibration method. With the method, a technical problem of high complexity of the existing multi-camera calibration method can be solved. According to the technical scheme, camera attitudes are obtained in real time by using an ORB-SLAM technology, so that dependence on a specific hardware platform is avoided; and on the basis of same scene pictures collected by multiple cameras at different times, resolving of a position relationship among the multiple cameras is carried out. The method is designed simply; overlapped-view-free or low-overlapped-view calibration of multiple cameras can be realized rapidly and accurately.

The invention discloses a color-coding-based multi-view dimensional stereoscopic-vision foot type three-dimensional measuring method and system and a medium. The method comprises: with a multi-view image collection platform, multi-camera synchronous shooting of a foot type with a color coding pattern sock is carried out and image acquisition is carried out; according to the collected image, internal and external parameters of the cameras are obtained based on a spherical multi-camera calibration algorithm; on the basis of internal and external parameters of the cameras, foot type dense featurepoint matching is carried out; and model construction and foot type parameter measurement are carried out based on the internal and external parameters of the cameras and the foot type dense featurepoint matching result. The color-coding-based multi-view dimensional stereoscopic-vision foot type three-dimensional measuring method and system have advantages of low cost, great easiness in featurepoint identification and matching, high generated point cloud density, and fast computing speed and the like. And the foot model and related size parameters are accurately and completely obtained.

The invention protects a multi-camera calibration method based on the minimum optical path principle and a transparent glass calibration plate and is applied to the fields of visual detection and image measurement. According to the method, the transparent glass plate is utilized to complete calibration of multiple cameras. The whole multi-camera calibration process comprises a first stage, whereinthe visual field of each camera is carefully adjusted to meet the requirement that identical covered regions exist between every two cameras; a second stage, wherein a Zhang Zhengyou calibration method is utilized to obtain internal parameters of the cameras; a third stage, wherein a pin-hole imaging model and a refraction projection model are used to acquire external parameters of the cameras infront of the calibration plate and external parameters of the cameras behind the calibration plate respectively; and a fourth stage, wherein the external parameters of all the cameras are unified under the same coordinate system. Compared with the prior art, by the adoption of the multi-camera calibration method based on the minimum optical path and the transparent glass calibration plate, calibration of multiple cameras in the system can be completed at the same time, and accuracy and reliability are improved.

The invention discloses a camera parameter obtaining method of a vehicle-mounted multi-camera surround view system, which comprises the following steps: obtaining a calibration object image collectedby a target camera, the target camera being any one of a plurality of cameras, a preset number of calibration objects being displayed in the calibration object image, and the calibration objects beingchessboards; extracting feature points in the calibration object graph, wherein the feature points are chessboard corner points in the chessboard; obtaining a first coordinate of the feature point onthe calibration object image and a second coordinate of a checkerboard corner point corresponding to the feature point on a checkerboard coordinate system; and inputting the first coordinates and thesecond coordinates into a preset machine model to obtain internal parameters and external parameters of the target camera. The technical problems of low efficiency and low automation degree of a multi-camera calibration process of a vehicle-mounted all-round view system in the prior art are solved. Meanwhile, the invention further discloses a camera parameter obtaining device of the vehicle-mounted multi-camera surround view system.

The invention discloses a multi-camera calibration method for 3D panoramic parking. According to the method, a 3D calibration object is placed in the angle-of-view areas of cameras sequentially according to preset distance and angle to obtain images, and the corresponding relationship between the pixels of the input images of the cameras and the actual environment is worked out; or, the 3D calibration object is placed in two overlapping areas to obtain a primary image, and the 3D calibration object is rotated by a fixed angle to obtain a secondary image. According to the invention, calibration of four cameras can be completed by placing the 3D calibration object four times or placing the 3D calibration object twice and rotating the 3D calibration object twice. An image spliced after calibration is a 3D image. The visual range is increased effectively, the image is not distorted, and users do not feel dizzy when viewing the image. The method of the invention is simple and convenient in operation. With the method, multiple cameras can be calibrated quickly, and convenience is brought to users.

The embodiment of the invention discloses a multi-camera calibration method and device based on a reference object. The multi-camera calibration method is used for solving the technical problems that existing calibration methods are not high in calibration precision, the positional relation between cameras is limited, and the existing methods are complicated. According to the embodiment, the multi-camera calibration method based on the reference object comprises the steps that S1, information of measurement performed on the preset reference object through multiple cameras is received, and feature points with a large surface curvature change of the reference object are acquired; S2, the positional relation between the cameras in an arbitrary number is obtained through the feature points; and S3, a global coordinate system with any feature point on the reference object being the origin or based on a local coordinate system corresponding to any camera is obtained through the positional relation between the cameras in the arbitrary number.

The invention discloses a product morphology detecting method based on Kinect. According to the method of the invention, three-dimensional reconstruction is performed on a product in a streamline by means of Kinect deep sensing equipment, and then the reconstructed model is compared with a preset product model for determining whether the product is a qualified method. The product morphology detecting method is advantageous in that a new improvement method is presented for multi-camera calibration, namely a cylinder manner is utilized for calibration; and a rotation translation matrix of the camera can be obtained accurately in higher speed. Compared with a method for performing integral filtering processing on a depth image for eliminating holes, the product morphology detecting method isadvantageous in that a target area is preliminarily selected by means of a rectangular block; and then an improved K-means clustering method is utilized for sufficiently eliminating product backgroundnoise and effectively extracting the product morphology. Compared with a traditional product detecting method, the method according to the invention has advantages of easy configuration, simple structure, high practicability, high speed and remarkable enterprise production cost reduction. Furthermore an error detecting rate of the product is reduced to a certain extent.

The invention relates to the field of video monitoring and discloses a visualization method of multi-camera calibration results and a system thereof. According to the invention, the method comprises following steps of acquiring a plane mapping relation between a first camera image and a second camera image; receiving input information from an input device; according to the input information, generating and displaying a first auxiliary line on the first camera image; acquiring at least two points on the first auxiliary line, using the at least two points as a first point set and mapping all points in the first point set to the second camera image according to the plane mapping relation to serve as a second point set; and according to all points in the second point set, displaying a second auxiliary line on the second camera image. According to the invention, results of the multi-camera calibration are described through the auxiliary lines, thereby facilitating rapid and direct verification and observation of calibration conditions and confirmation of calibration precision.

A chart for calibrating a system of multiple cameras, the chart comprising: a background; an array of dots contrasting the background, wherein the array of dots are arranged in rows and columns, wherein the array of dots comprise a first dot array, and a second dot array, wherein the first dot array fully occupies a first region of evenly spaced dots with a first dot density, the second dot array fully occupies a second region of evenly spaced dots with a second dot density, and wherein the second region is enclosed within the first region; a group of first markers in the first region, a group of second markers in the second region, and a third marker at the center of the chart, wherein each second marker is closer to the third marker than each first marker.

The invention discloses an identification system and method of a panoramic looking-around multi-camera calibration rod, and belongs to the fields of image measurement and computer vision. According to the identification system and method, a panoramic looking-around multi-camera calibration system based on the calibration rod with six table tennis lampshade LED lamps is established, a multi-camera 3D identification algorithm of the novel calibrating object is provided, eight cameras are adopted for 360-degreee all-direction shooting, calibration image identification respectively under 3 different environments, such as daytime outdoor, indoor light-on and indoor light-off conditions, is realized on Visual C and OpenCV platforms, and a collinear theory and a least square method are adopted for rejecting ineffective points, so that the identification precision and effect are substantially improved, adaptive tracking of a moving object under different environment is realized, and the calibration accuracy, robustness and flexibility of the cameras are effectively improved.

The invention relates to the field of three-dimensional measurement, in particular to a large-view-field stereoscopic vision calibration method for high-precision calibration of a stereoscopic vision system during large-size measurement, which comprises the following steps of: establishing a relationship between a camera model coordinate system and a pixel coordinate system; image distortion correction is carried out; measuring to obtain three-dimensional space coordinates; calculating a rough pose relation of the image detection; solving a matrix of a pose relation; calculating an essential matrix, and establishing a multi-camera calibration model; calibration of parameters in the camera is realized; calculating theoretical coordinate values and measurement coordinate values of three points which are not on the same straight line; and obtaining an optimal solution of the rotation matrix and the translation vector. Finishing the correction of the scale factor; and an optimal scale factor is obtained. According to the invention, the monocular camera is creatively utilized to calibrate the coding points and the mark points in the field of view to be measured, and on the basis of the obtained three-dimensional coordinate values, the attitude relationship among different images is estimated by utilizing an eight-point algorithm, so that high-precision and high-efficiency calibration of the stereoscopic vision measurement system is realized.

The invention provides a multi-camera calibration method and device, a storage medium and an electronic device, and the method comprises the steps: obtaining an image, obtained through the simultaneous photographing of a calibration object through all cameras, and the calibration object comprising a target feature point; determining a target feature point corresponding to the image coordinate datain each image; and according to the image coordinate data and the space coordinate data of the target feature point, determining parameters of a camera for calibration, the space coordinate data being coordinate data of an actual space position in a world coordinate system. According to the invention, the joint calibration of the plurality of cameras can be efficiently completed in a manner thatthe plurality of cameras shoot only one image at the same time, the calibration time of the plurality of cameras is effectively shortened, the calibration accuracy is ensured, and the calibration efficiency is improved.

The invention provides a multi-camera calibration method and device, a storage medium and electronic equipment. The electronic equipment comprises a plurality of cameras, the plurality of cameras arean ultra-wide-angle camera, a wide-angle camera and a long-focus camera. The method comprises the following steps: based on different angles, obtaining images obtained by simultaneously shooting a calibration object by each camera, wherein the calibration object comprises a target feature point; determining a target feature point corresponding to the image coordinate data in each image; and according to the image coordinate data and the space coordinate data of the target feature point, determining parameters of a camera for calibration, the space coordinate data being coordinate data of an actual space position in a world coordinate system. The plurality of cameras can calibrate the calibration object at the same time, the calibration angles are multiple, the calibration execution time isshortened, the calibration errors are reduced, and the calibration efficiency is improved.

Provided is a multi-camera coordination control and quick precise target identification device. Eight cameras are used for observing a view field by 360 degrees, a matching result of a blurred image and a clear image collected before and after focusing and a geometric optical path are used for quickly and accurately positioning an object, and the object is identified and positioned through image processing. The eight cameras all have an automatic focusing function, after the target is identified, the corresponding cameras are triggered to collect two images before and after focusing, and a matching algorithm in a deigned embedding system and an optical equation are used for calculating the position of the target. An intelligent robot which is provided with a jumping spider imitation visual system has a detection view field close to 360 degrees, and complex calculation of coordinate transformation of a multi-camera calibration coordinate system is avoided through the special target positioning method.

A novel multiple camera calibration algorithm uses human joint points for matched key points. A recent machine-learning based human joint detector provides joint positions with labels (e.g. left wrist, right knee, and others). In single person situation, it directly provides matched key points between multiple cameras. Thus, the algorithm does not suffer a key-point matching problem, even in a very sparse camera configuration, which is challenging in the traditional image feature-based method. This algorithm provides easy setup for a multiple camera configuration for marker-less pose estimation.