72 results about "Camera matrix" patented technology

Aspects of the present invention include systems and methods for recalibrating projector-camera systems. In embodiments, systems and methods are able to recalibrate automatically the projector with arbitrary intrinsic and pose, as well as render for arbitrarily desired viewing point. In contrast to previous methods, the methods disclosed herein use the observing camera and the projector to form a stereo pair. Structured light is used to perform pixel-level fine reconstruction of the display surface. In embodiments, the geometric warping is implemented as a direct texture mapping problem. As a result, re-calibration of the projector movement is performed by simply computing the new projection matrix and setting it as a camera matrix. For re-calibrating the new view point, the texture mapping is modified according to the new camera matrix.

A three-dimensional (3-D) shape measurement method using a Moire measurement principle and a Stereo vision measurement principle is provided. The method comprises; a first step to detect candidate points and 3D positions of the candidate points in world coordinates which are identical results from typical Moiré technique by using a pattern projector, which adjusts a pitch of a fringe pattern and projects the fringe pattern to a measurement object, and a first camera, which detects a modulated fringe pattern caused by shape of shape of the object, by means of obtaining candidate points by comparing a reference fringe pattern to the modulated fringe pattern and calculating 3D positions at an arbitrary point on an image captured from the first camera; a second step to determine a final matching point among the candidate points detected from the first step by using the Stereo vision technique with the first and the second cameras, by means of projecting the 3D positions of the candidate points to an image plane of the second camera based on a camera matrix of the second camera, comparing intensity value of a projected point in the image plane of the second camera to intensity value of given point in the image plane of the first camera, and judging whether the candidate point is matched with an arbitrary point by using the stereo vision criteria; and a third step to measure the depth value of the arbitrary point by using the resultant value of the final matching point, which was determined at the second step. Therefore, the error of 2π ambiguity of the typical Moire principle is eliminated, and thus 3-D shape information can be more rapidly and accurately measured.

The invention discloses a synthetic method of stereoscopic elements in a combined stereoscopic image system collected by a sparse lens, and belongs to the technical field for generating stereoscopic images. The method comprises the following steps of: arranging a sparse camera matrix; calculating optical parallax image between each shot image and an adjacent horizontal image and between each shotimage and an adjacent vertical image; expanding a binocular stereoscopic vision model into a multi-eye stereoscopic vision model, and calculating spatial location coordinates of actual object points corresponding to the pixel points in each image according to a horizontal parallax image and a vertical parallax image of each image; transforming the spatial location coordinates of the actual objectpoints corresponding to the pixel points calculated in all images into the same coordinate system; virtually simulating a camera array, projecting the actual object points corresponding to the pixel points in each image under the same coordinate system to each virtual camera to generate sub-images; interpolating the points without being filled; and generating a stereoscopic element image. The synthetic method can be used for accomplishing the parallel generation of a great quantity of virtual view points and is more suitable for virtual synthetics of the stereoscopic element image; and therefore, for an image source of the combined stereoscopic image system, live action shooting can be carried out in place of computer production.

The invention, in view of the disadvantages in the prior art that a camera matrix for a three-dimensional modeling image capturing device has high cost and a photo taken by a low-cost image capturingdevice is liable to produce a disordered light and shadow effect, provides a three-dimensional modeling image capturing device for a high reflective product and belongs to the technical field of three-dimensional modeling photography. The three-dimensional modeling image capturing device includes a platform for holding a photographed object, a camera for photographing the photographed object, a driving device for driving the camera to perform a circular motion around the platform, and a soft light cover for filtering light. The driving device is provided with an adjusting device for adjustingthe height of the camera. The platform, the camera, the driving device and the adjusting device are all disposed in the soft light cover. The device only needs one camera and is low in cost. The relative position between the object to be photographed and a light source does not change during image capture, so the disordered light and shadow effect cannot be produced, and an obtained photograph hasa good effect.

The present invention provides a 3D four-dimensional data collection system and method based on a visible light camera matrix. The system comprises: a support structure, an arc bearing structure arranged on the support structure, a plurality of cameras arranged on the arc bearing structure and forming a camera matrix, and a data processing unit connected with the cameras. The cameras are configured to perform collection of iris feature information to obtain multiple iris images and transmit the multiple iris images to the data processing unit; and the data processing unit is configured to perform processing of the iris images. The collection method comprises the steps of: obtaining iris images through cameras; transmitting the iris images obtained by the cameras to a central processing device for processing and calculation; and generating iris 3D four-dimensional data of a collection object. According to the embodiment of the invention, the multi-camera control technology is employed to perform collection of the iris feature information so as to observably improve the collection efficiency of the iris feature information.

The invention relates to a three-dimensional target detection method and system based on image restoration. The method comprises the following steps: obtaining an RGB image and radar point cloud of athree-dimensional target; generating a two-dimensional target detection box on the RGB image according to a two-dimensional target detection algorithm; for the picture with the shielded target, carrying out instance segmentation by adopting an instance segmentation algorithm to obtain a mask at the shielded position of the target, and then calculating a complete mask of the target according to morphological closed operation; converting the radar point cloud into a depth map through a camera matrix, performing image restoration on the shielded part of the target on the depth map, and extractingdepth information of the target in a depth map form according to a complete mask of the target after the restoration is completed; converting the depth information in the depth map form of the targetinto restored point cloud; and inputting the repaired point cloud into a three-dimensional target detection network for three-dimensional target detection. Compared with the prior art, the three-dimensional target detection method plays a role in reducing offset and improving precision in three-dimensional target detection.

The invention provides a glasses automatic design method based on 3D head data. According to the glasses automatic design method based on 3D head data a camera matrix is adopted for acquisition of head characteristic information so as to obtain a plurality of head images, characteristic points from the plurality of head images are extracted, and head characteristic point cloud data is generated; ahead 3D model is constructed based on the point cloud data to realize acquisition of 3D head data; glasses frame design parameters are extracted based on the 3D head data, and a glasses frame 3D model is formed; glasses lens parameters are determined based on the eyesight situations of the two eyes and the glasses frame design parameters, a glasses 3D model formed based on the glasses frame modeland the glasses lens parameters is displayed, and parameters are adjusted based on the display effect to complete glasses design. The glasses automatic design method is capable of realizing customized automatic design and manufacturing of glasses; visible light camera matrix 3D scanning equipment control technology is adopted for head characteristic information collection, and acquisition efficiency and design accuracy can be increased obviously.

The invention relates to a camera self-calibration method based on three orthogonal direction end points. Contact among the three orthogonal direction end points, a camera matrix and an inner parameter matrix is deduced based on the three orthogonal direction end points. Aiming at five parameter camera models, the camera self-calibration method based on the three orthogonal direction end points is provided. The method comprises the following steps of solving coordinates of the three orthogonal direction end points; solving proportional factors according to the contact of the end points and the camera matrix, building a restriction equation of the inner parameters of the camera, and linearly solving five inner parameters of the camera. The whole solution of the inner parameters of the camera can be realized by only needing two views; and by the calibration method, the automatic calibration can be realized, and faults caused by measurement during calibration are reduced.

The invention provides a method for registering a camera spindle and a laser beam in parallel. The method can be used in a system including a camera and a laser transmitter (a laser rangefinder or a laser structure light) to register the system in parallel; namely, the placing positions of the camera and the laser transmitter are adjusted so that the camera spindle of the camera is parallel to the laser beam transmitted by the laser transmitter. The method comprises the following steps of: shooting images of a chessboard calibration block and light spots, wherein the light spots are formed by the laser beam on the chessboard calibration block; and then calibrating camera parameters (a camera matrix) by using shot images and estimating the camera spindle; next, calibrating a laser beam origin by using a plane homographic principle; and finally, estimating an image point parallel to the laser beam by using the plane homographic principle and calibrating the laser beam. By using the method for registering the camera spindle and the laser beam in parallel, provided by the invention, the calibration can be finished only by shooting four images and only one chessboard calibration block is needed for assistance; therefore, the method for registering the camera spindle and the laser beam in parallel, provided by the invention, has the advantages of high calibrating speed, easiness for realization and higher accuracy.

The embodiment of the invention provides a roadside parking management system based on a camera matrix. The roadside parking management system based on the camera matrix comprises a roadside parking access region camera, a first camera matrix, a main controller and at least one installation rod or installation position; the roadside parking access region camera is used for acquiring regional images of a roadside parking entrance region and a roadside parking exit region; the first camera matrix is used for acquiring a vehicle information image of a berth in the first predetermined acquisitionrange from the installation rod or the installation position; and the main controller is used for determining the occupation information of each berth in the roadside parking region according to the acquired regional image of the roadside parking entrance and/or exit region and the vehicle information image of the berth in the first predetermined acquisition range, so that roadside parking management is realized. By means of the embodiment, unattended management on roadside parking is realized; and a vehicle parking event can be judged in time more accurately.

The invention discloses a truck carriage volume measurement method based on a single image. The defects that an existing photoelectric volume measurement system can only be applied to a detection channel with the fixed position, and multiple laser devices, cameras and sensors are needed are mainly overcome. The truck carriage volume measurement method includes the steps of 1, inputting the image of a truck, and calibrating an intrinsic parameter matrix of a camera through end points in the image of the truck in three orthogonal directions; 2, calculating normal vectors of wheels and coordinates of circle centers of the wheels in cooperation with the intrinsic parameter matrix and an elliptic equation of the wheels of the truck, and obtaining a carriage plane equation; 3, calculating extrinsic parameters of the camera through the end points and the coordinates of the circle centers of the wheels; 4, solving space coordinates of top points of a carriage in cooperation with the carriage plane equation and a camera matrix operation equation, and the solving the volume of the carriage. By means of the truck carriage volume measurement method, the volume of the carriage of the truck can be rapidly and accurately measured, the automation degree is high, and the truck carriage volume measurement method can be used for a charging system of an expressway toll-gate.

The invention provides a measuring device based on a virtual matrix and a related device, wherein, the measuring device comprises an image acquisition device, which is used for providing an acquisition area and acquiring corresponding images; a collection-area move device, wherein the image acquisition device is used for driving the acquisition area of the image acquisition device to move to different positions to form other acquisition areas, so that a virtual image acquisition matrix composed of a plurality of acquisition areas is formed in a certain time space, and the image acquisition device obtains images of different directions of the target object from the plurality of acquisition areas; the image acquisition device is used for driving the acquisition areas of the image acquisitiondevice to move to different positions to form other acquisition areas; an image processing device that processes the image obtained by the virtual image acquisition matrix to obtain 3D information ofthe object; a measuring device measuring a geometric size of a target object according to 3D information of the target object. For the first time, the problem of low resolution of multi-camera matrixacquisition due to the camera volume is noticed and put forward. It is proposed that the resolution can reach pixel level by forming virtual camera matrix in a certain period of time to improve the acquisition resolution.

The invention relates to a novel camera calibration method based on two crossed straight lines. The method comprises the following steps: (1), manufacturing a calibration object; (2), shooting and imaging; (3), detecting the coordinate of a pixel; (4), building an equation for each picture; (5), constructing a linear system of equations according to equations of all pictures shot in the step (2) to obtain the equations in the step (4), and solving the linear system of equations to obtain the a camera parameter matrix; and (6), performing non-linear optimization on the result in the step (5), constructing a target function, and using the minimum target function to obtain the minimum distance between the estimated image position of the camera matrix and the actual image position. According to the invention, the constraints of difficulty in free movement of the calibration object in the shooting process are eliminated, so as to achieve better calibration accuracy; and particularly, a pre-treatment technique provided by the invention in the calibration process is beneficial to improving the accuracy and the stability of the calibration algorithm through maintaining the linear relation of the calibration object to reduce the coordinate error of an image point.

The invention discloses a pipeline measurement method based on multi-view vision. The method is composed of four parts of image acquisition and preprocessing, center line and outer diameter extraction, camera calibration and camera matrix conversion, pipeline center line matching and outer diameter expansion. A camera is placed on a built measurement frame to carry out image acquisition and preprocessing and acquire a pipeline gray scale picture. The method comprises the following steps of acquiring a pipeline center line by applying distance transformation and a minimum path method, extracting feature points of the pipeline center line, and discretizing the feature points; calibrating the corresponding binocular cameras (a front side camera and a rear side camera), and obtaining a conversion matrix from a rear side camera coordinate system to a front side camera coordinate system; and recovering three-dimensional information of the pipeline center line by adopting a dynamic matchingmethod according to the obtained center line feature points and the camera calibration parameters, expanding outer diameter information obtained by processing pictures shot from different viewing angles on the three-dimensional pipeline center line, and finally obtaining the spatial position of the pipeline center line and the corresponding outer diameter information.

The invention provides an incremental multi-view three-dimensional reconstruction method, belongs to the technical field of computer vision, and solves the problem of utilization of reconstructed model information in a multi-view three-dimensional reconstruction process. The method comprises the following steps: 1, obtaining a three-dimensional reconstruction model by utilizing a motion recovery structure method, and respectively storing an image path, an image camera matrix and reconstructed three-dimensional point information into a memory; 2, reading a new image and a three-dimensional reconstruction model, determining an initial image pair according to an image feature matching relation and a two-dimensional three-dimensional mapping relation, and reconstructing a three-dimensional point cloud; 3, newly adding one to-be-reconstructed image each time, and reconstructing the to-be-reconstructed newly added image until all the images are reconstructed; and 4, carrying out multi-view stereoscopic vision dense reconstruction based on a patch on the reconstructed sparse point cloud model to complete multi-view three-dimensional reconstruction. The method is high in applicability, andthe processes of manual image set screening and sorting and three-dimensional registration are simplified.

The invention discloses a game map mirror symmetry method which comprises the steps of before a camera module renders, turning over the position of a camera to the left lower corner of a game map, multiplying a world transfer camera matrix by a mirror image matrix, and starting reversal removal of a bottom-layer graphical interface; rendering through the camera module, resetting the world transfercamera matrix after accomplishing rendering, resetting the position of the camera, and closing the reversal removal; modifying an X-axis forward direction of a virtual joystick into an X-axis reversedirection corresponding to a game character; and modifying a Y-axis forward direction of the virtual joystick into a Z-axis reverse direction corresponding to the game character. In addition, the invention further discloses a game map mirror symmetry system. By adopting the game map mirror symmetry method and the game map mirror symmetry system provided by the invention, the visual angles of bothgame parties can be completely ensured to be consistent, and the game balance problem brought by a visual angle difference cannot occur, and the problem that the game players disagree path selectionof a game map cannot exist.

The invention discloses a vehicle-mounted electronic rear-view mirror system, which comprises a camera matrix, at least two first independent display screens and a video matrix switching module. The camera matrix comprises at least two cameras; The video matrix switching module is connected with the camera matrix and the at least two first independent display screens. The video matrix switching module receives the video signal inputted from the camera matrix and transmits the video signal to the first independent display screen for display. The video signals displayed on different first independent display screens are collected by different cameras, and each of the first independent display screens only displays the video signals collected by one camera. The vehicle-mounted electronic rearview mirror system provided by the embodiment of the invention ensures that the electronic rearview mirror system can be started quickly, can acquire the information around the vehicle in time, can assist the driver in observing the blind area vision, effectively eliminates the hidden danger, and improves the safety of driving.

The invention discloses an optimization method of SIFT characteristic matching points based on limit restraint, and aims at overcoming disadvantages of SIFT characteristic matching points. Obtained matching points at present are divided into a high-quality point set V11, a standard point set V12 and a to-be-measured point set V13, the standard point set V12 and to-be-measured point set V13 are updated, the high-quality point set V11, the updated standard point set V12 and the updated to-be-measured point set V13 are combined, a basic camera matrix F is calculated again in an RANSAC method, and exterior points are rejected. Thus, the matching rate is effectively improved, the computational complexity in the budgeting process is low, and computation is rapid.

The invention provides a 3D four-dimensional hand image data recognition method and device. The method comprises the steps that the acquisition of the four-dimensional data of a hand is completed by using a camera matrix, and a four-dimensional model of the hand features is constructed according to a plurality of hand feature images, so as to realize the collection of the four-dimensional data ofthe hand features of a person; the identity information of a character is used as an identification mark to form a database comprising a plurality of pieces of four-dimensional data of the hand features; the four-dimensional data of the hand features stored in the database is searched through the identity information of a target character, and using the identity information of the target person tofind the four-dimensional data of the hand features stored in the database, and a sky-point point cloud comparison method is used for recognizing the identity of the target character. The invention also provides a 3D four-dimensional hand image data recognition device. The device comprises a hand feature information collection device, a hand feature four-dimensional data storage device, and a target character identity recognition device. The sky-point point cloud comparison method based on the spatial domain direct matching is used for the fitting of features points, thereby achieving the quick fitting comparison of the feature points of the hand, and achieving the quick authentication and recognition of an identity.

The present invention provides a method and a device for biological feature 3D data collection based on a visible light camera. The method comprises the steps of: employing a camera matrix formed by aplurality of visible light cameras to perform collection of biological feature information, and obtaining multiple biological feature images; processing the biological feature images, and extractingfeature points of the biological feature images; based on the extracted feature points in the biological feature images, generating feature point cloud data of the biological feature features; and constructing a 3D model of the biological features according to the feature point cloud data to achieve collection of the biological feature 3D data. Through adoption of the camera matrix control technology, the method and the device perform collection of the biological feature information to observably improve the collection efficiency of the biological feature information; and moreover, the methodand the device can employ the collected feature information of the biological features on a space to completely restore the features of the biological features on the space so as to provide infinite possibility for follow-up application of the biological feature data.

The invention provides a 5G-based cloth defect detection method. The method comprises the following steps of: configuring the cloth winding speed to be consistent with the movement speed of an annularcirculating operation line fixedly connected with a plurality of industrial camera groups; waiting for an image acquisition instruction by the industrial camera; sending the image information of thecurrently detected cloth shot by the plurality of industrial camera groups to a cloud processor through 5G communication; and carrying out defect detection on the image information by the cloud processor, marking the images with defects, sorting the marked images according to a time sequence, and splicing the sorted images to obtain a full image with defect marks of the currently detected cloth. The invention further relates to a 5G-based cloth defect detection system. According to the method, image acquisition is carried out by adopting a circulating camera matrix, and the image is kept at the same speed as the cloth, so that the high definition and stability of the image are ensured; and meanwhile, the 5G technology and cloud detection are utilized to realize rapid and efficient defect recognition and accurate positioning of defects, so that rapid searching and repairing of the defects are facilitated.

The embodiment of the invention provides an information acquisition method, device and system, and the method comprises the steps: adjusting the shooting parameters of at least two cameras contained in a camera matrix to be target shooting parameters according to a shooting parameter adjustment signal, and enabling at least two cameras to be arranged in a preset mode; controlling at least two cameras to synchronously acquire target information at the shooting moment according to the synchronization signal, wherein the target information comprises image information shot through the target shooting parameter and audio information collected at the shooting moment, and the synchronization signal is corresponding to at least one shooting moment; and for each shooting moment, synchronously acquiring the target information acquired by at least two cameras. According to the invention, the consistency of the shooting parameters of the camera can be maintained to the maximum extent, the shootingcontinuity can be ensured when the environment state changes, and the consistency of the synchronous shooting effect and the picture can be ensured.

The invention relates to an inclined image relative orientation method based on essential matrix decomposition. The inclined image relative orientation method is characterized by comprising: calculating an essential matrix E; recovering a camera matrix P from the essential matrix; converting the camera matrix P into consecutive relative orientation element initial values; converting the consecutive relative orientation element initial values into independent orientation element initial values; and constructing an independent relative orientation error equation, solving the independent relativeorientation element correction value through a least square method, updating the independent relative orientation element, and carrying out least square iteration solving so as to achieve the inclined image relative orientation function. According to the present invention, with the relative orientation method comprising obtaining the consecutive relative orientation element initial values based on the essential matrix decomposition and performing using least square iteration optimizing, the problem that the traditional relative orientation method cannot be adapted to the inclined image relative orientation is solved, and the precision requirement of the photogrammetry on the relative orientation is met; and the method of the present invention is suitable for the inclined aerial photography images with large relative attitude and any relative positions.

The invention discloses a camera calibration method and a terminal. The method comprises the steps of obtaining historical image information of a camera; establishing an image coordinate system according to the historical image information; respectively acquiring image coordinates of the main point and the three vanishing points in the image coordinate system; performing calculating to obtain an internal parameter matrix of the camera according to the focal length of the camera and the image coordinates of the main point; performing calculating to obtain a camera matrix according to the internal parameter matrix and the image coordinates of the three vanishing points; selecting two non-parallel reference line segments according to the historical image information; respectively obtaining end point image coordinates and real length information of the two reference line segments; performing calculating to obtain a first scaling factor and a second scaling factor according to the end pointimage coordinates and the real length information; and performing calculating to obtain a calibration matrix of the camera according to the camera matrix, the first scaling factor and the second scaling factor. The calibration matrix is more accurate through calculation of the double scaling factors, and real information in a camera image can be accurately calculated.