25 results about "Omnidirectional camera" patented technology

A camera system includes a housing and camera modules (e.g., digital cameras) that are aligned with fields of view that extend below the housing. The camera modules are provided about a perimeter of the housing and mounted to a bench within the housing. Two camera modules have axes of orientation extending below and away from a centroid of the camera system. Two camera modules have axes of orientation extending below and toward the centroid of the camera system. The housing includes inlets and outlets that enable air to flow past the camera modules and other components within the housing. Images captured by the camera modules of the camera system may be utilized for any purpose.

The invention provides a method and a system for self-calibrating an omnidirectional radiation field. The method comprises the following steps: generating a 3D Gaussian as an initial representation of an omnidirectional radiation field based on an acquired 360-degree image; rasterizing the 3D gauss into a 2D omnidirectional image, the rasterizing comprising sputtering the 3D gauss onto a unit spherical surface and expanding the 3D gauss into an equirectangular projected 2D omnidirectional image; the 360-degree image is resampled based on an omnidirectional camera model, an undistorted image is obtained, and the omnidirectional camera model is constructed based on a fixed spherical grid structure and a trainable focal length coefficient and an angle distortion coefficient; the 3D Gaussian, camera pose, and the omnidirectional camera model are jointly optimized by minimizing the loss of luminosity between the 2D omnidirectional image and the undistorted image.

The invention discloses a 3D target visibility processing method, system and device of a panoramic camera and a medium, and relates to the technical field of image processing. Target voxel data and a 3D target detection frame are acquired; selecting a target 3D frame point in the 3D target detection frame, and performing position transformation on the target 3D frame point and the target voxel data to obtain target positions of the target 3D frame point and the target voxel data under each panoramic camera; based on the target position and the projection of the voxel data in the corresponding panoramic camera image, generating a theoretical projection region area and an effective projection region area of the 3D target detection frame in the panoramic camera; and performing weighted fusion on the theoretical projection region area and the effective projection region area, generating a fusion visible proportion of the target object according to a weighted fusion result, and generating a target visibility result of the target object in the panoramic camera according to the fusion visible proportion. The fusion visible proportion is matched with the complete size of the target object and the image constraint of the panoramic camera, and the method is accurate and high in efficiency.

The invention discloses an omnidirectional structured light 3D sensor calibration method and system based on model-free learning, and the method comprises the steps: synchronously collecting image data of an omnidirectional camera and a measurement sensor, carrying out the multi-camera coordinate alignment of the measurement sensor, and building a unified reference coordinate system; in a unified reference coordinate system, light plane calibration is carried out, and a light plane equation in a measurement sensor coordinate system is obtained; based on the light plane equation, a 2D-3D corresponding data set between the omnidirectional camera and the measurement sensor is constructed, and the data set comprises a sample pair composed of two-dimensional pixel points in an omnidirectional camera image and three-dimensional coordinate points corresponding to the two-dimensional pixel points in a measurement sensor coordinate system; and based on the 2D-3D corresponding data set, training an end-to-end deep learning model, and learning a direct mapping relation from the image coordinates of the omnidirectional camera to the 3D coordinates of the measurement sensor to complete calibration. According to the invention, a general calibration framework without an explicit physical model is realized.

Methods and processors for rendering a 3D Gaussian are disclosed. The processor is configured to acquire the 3D Gaussian to be rendered, and a camera model representing an omnidirectional camera with an optical axis, updating a color of the 3D Gaussian using spherical harmonics, update a position of the 3D Gaussian by moving the 3D Gaussian towards the optical axis, update a scale of the 3D Gaussian by compressing the 3D Gaussian in at least one of a tangential direction and a polar direction relative to the optical axis, and render an updated 3D Gaussian onto a 2D plane using a 3DGS model.

A method includes: pairing a head-mounted virtual reality device with an omnidirectional camera; displaying the converted media file associated with media content corresponding to the external environment relative to the vehicle to the user; generating one or more augmented reality images; and superimposing one or more augmented reality images onto the displayed converted media file.

To simplify the correction of a coordinate system and scale in three-dimensional data processing using an omnidirectional image captured by an omnidirectional camera.SOLUTION: A three-dimensional data processing system 1 is configured to: estimate the position and orientation of an omnidirectional camera 5 during capture from the correspondence between feature points and omnidirectional images; restore first-density point cloud data indicative of three-dimensional distribution of first-density feature points on the basis of the feature points and the position and orientation of the omnidirectional camera 5 during capture; perform a correction to align the coordinate system and scale of the first-density point cloud data with the absolute coordinate system and absolute scale of the real physical space 40 in which the images were captured by the omnidirectional camera 5 on the basis of at least one of the position and orientation of the omnidirectional camera 5 during capture and the information related to the feature points; and restore second-density point cloud data indicative of the three-dimensional distribution of feature points at a second density higher than the first density from the correspondence between the corrected first-density point cloud data and the omnidirectional images.SELECTED DRAWING: Figure 5

Methods and processors for rendering a 3D Gaussian are disclosed. The processor is configured to acquire the 3D Gaussian to be rendered, and a camera model representing an omnidirectional camera with an optical axis, updating a color of the 3D Gaussian using spherical harmonics, update a position of the 3D Gaussian by moving the 3D Gaussian towards the optical axis, update a scale of the 3D Gaussian by compressing the 3D Gaussian in at least one of a tangential direction and a polar direction relative to the optical axis, and render an updated 3D Gaussian onto a 2D plane using a 3DGS model.

One of a left side bin and a right side bin is provided with a driving sliding table, the other one of the left side bin and the right side bin is provided with a follow-up sliding table, a first moving block of the driving sliding table and a second moving block of the follow-up sliding table are fixedly connected to the two sides of a middle bin respectively, and the driving sliding table drives the first moving block to move up and down; the first moving block drives the middle bin and a second moving block of the follow-up sliding table to move up and down. A pitching rotation driving structure is arranged in the middle bin, the two sides of the lens bin are rotatably connected to the inner sides of the two sides of the middle bin respectively, and the pitching rotation driving structure drives the lens bin to rotate in a pitching mode; the left side bin and the right side bin are fixedly connected with the rotating base; and a horizontal rotation driving structure is arranged on the fixed base, is connected with the rotating base and is used for driving the rotating base to rotate horizontally. According to the invention, active compensation can be carried out on the large longitudinal amplitude generated by the holder in the vehicle movement process, and dead-corner-free stable acquisition of videos is realized.

This application provides a omnidirectional camera, comprising: a base, the first surface of which is spherical; a spherical camera detachably mounted on the first surface via a mounting surface, the spherical camera being configured to move in any direction along the first surface of the base; the first surface of the base is provided with a plurality of electromagnet groups, each electromagnet group comprising a plurality of electromagnets, the spacing between adjacent electromagnets being a first distance; the mounting surface of the spherical camera is provided with a plurality of permanent magnet groups corresponding to the plurality of electromagnet groups, each permanent magnet group comprising a plurality of permanent magnets, the permanent magnets and the electromagnets having opposite polarities at their closest ends, the spacing between adjacent permanent magnets being a second distance, the second distance being unequal to the first distance. The omnidirectional camera of this application allows the camera to move in any direction within the base by controlling the attractive force between the camera and the base.

The utility model discloses a multifunctional road detection vehicle which comprises a vehicle body, one side of the vehicle body is fixedly connected with a fixing support, and the top of the fixing support is fixedly connected with a warning rod. The laser profiler at the bottom of the fixing support is used for detecting the flatness of a road surface in the running process of a vehicle, then the front road condition camera monitors the road condition in the running direction of the vehicle in real time, the omni-directional camera is responsible for monitoring the road condition around the vehicle, and then the satellite positioning system ensures the accuracy of the running route of the vehicle. Meanwhile, the three-dimensional pavement disease camera monitors a driving pavement, pavement damage and deterioration phenomena can be found in time and fed back and repaired in time, and then necessary illumination is provided through the auxiliary illuminating lamp when the laser ranging mechanism measures the pavement distance, so that accurate detection data in each distance are obtained; and the pavement structure depth tester is used for detecting the structure and the laying depth of the pavement, so that the efficiency and the accuracy of road detection are improved.

The invention discloses a calibration object-free space-time calibration method for a multi-sensor system, and relates to the technical field of sensor calibration, and the method comprises the steps: obtaining the external parameters and motion tracks of an inertial unit, a camera and a laser radar, and determining the dynamic constraint of the motion; projecting the three-dimensional laser radar edge points and the two-dimensional camera image edge points onto the same unit spherical surface in a unified manner, and further establishing static geometric constraints; and constructing a least square optimization problem according to the dynamic constraint and the static geometric constraint of the motion, and solving the least square optimization problem to obtain calibration values of all the to-be-calibrated parameters. According to the method, calibration objects are not needed, the omnidirectional camera is supported, efficient and accurate unified joint calibration can be carried out on the vision-inertia-laser radar, and an accurate sensing basis can be provided for the omnidirectional camera in a visual odometer and a mapping task.

The invention discloses a ship camera attitude correction method based on multi-camera cooperation, and relates to the technical field of ship navigation perception, and the method comprises the steps: carrying a plurality of cameras facing a bow, a stern, a port and a starboard in the front, rear, left and right directions of a compass deck; acquiring environment image sequences in front of the ship, behind the ship, in the left side direction and in the right side direction of the ship by using each carried camera; using a sea-sky-line identification method to obtain sea-sky-line characteristics of each frame of environment image; acquiring an inertial navigation data sequence by using a carried inertial measurement unit, and fusing the inertial navigation data sequence with the attitude correction sequence of the camera to obtain a camera attitude sequence at the current moment; and adjusting the camera attitude according to the camera attitude sequence. According to the method, dynamic, high-precision and omnidirectional camera attitude correction is realized based on a multi-camera cooperation mechanism, and the robustness of a ship sensing system is effectively improved.

Methods and processors for rendering a 3D Gaussian are disclosed. The processor is configured to acquire the 3D Gaussian to be rendered, and a camera model representing an omnidirectional camera with an optical axis, updating a color of the 3D Gaussian using spherical harmonics, update a position of the 3D Gaussian by moving the 3D Gaussian towards the optical axis, update a scale of the 3D Gaussian by compressing the 3D Gaussian in at least one of a tangential direction and a polar direction relative to the optical axis, and render an updated 3D Gaussian onto a 2D plane using a 3DGS model.

This invention relates to a device for multi-angle, omnidirectional camera installation and DIC measurement, comprising a frame, an observation platform inside the frame, an omnidirectional guide frame on the inner periphery, and a camera assembly and a progressive assembly ringed on the omnidirectional guide frame; the camera assembly is used for real-time measurement of the observed object, and the progressive assembly enables dynamic multi-angle adjustment of the camera assembly; the omnidirectional guide frame is formed by splicing multiple track arms end to end to form an annular disk, and progressive rings with semi-circular cross sections are evenly distributed on the surface of the track arms; the camera assembly includes a calibration camera and a follow-up camera; the follow-up camera is connected to the progressive assembly; the progressive assembly consists of a progressive hammer, with an umbrella-shaped locking rod on the outer periphery of the hammer and a T-shaped adaptive hammer core inside; air storage bladders are symmetrically wrapped on both sides of the hammer core, which are inflated or deflated through air inlet ports; a clamping spring is sandwiched between the air storage bladder and the inner wall of the hammer, which can reset as the air bladder changes, driving the follow-up camera to slide along the progressive ring to achieve dynamic adjustment, which eliminates the blind spot problem caused by the limited viewing angle in single-field measurement.

The invention discloses an intelligent inspection system for a coke oven, which is characterized in that a wall climbing trolley is arranged on the wall surface of the coke oven, and a reversing device is arranged at the wall corner; a routing inspection path is designed according to the wall surface around the coke oven, and path information is stored in a control system of the wall climbing trolley; the wall climbing trolley climbs along a preset path, and the universal camera and the sensor collect state information of the coke oven in real time; the data is transmitted to a remote monitoring center through the wireless communication module for real-time analysis; and when the wall climbing trolley arrives at the reversing device, the reversing device is started to guide the trolley to an adjacent wall surface. The wall climbing trolley can flexibly move along the wall of the coke oven, efficient inspection is achieved through the preset inspection path, the working efficiency is remarkably improved compared with manual inspection, and the wall climbing trolley is particularly suitable for continuous inspection tasks of large-scale coke ovens. Manual operation is replaced, and the safety of inspection work is improved.

The application discloses a kind of camera and vehicle extrinsic parameter online calibration optimization method and related equipment.Method includes: obtaining target image, extracting vanishing point in the target image, the target image is the image that omnidirectional camera is photographed when target vehicle walks in straight line;Obtain the first vanishing point and the second vanishing point satisfying the first condition, respectively join to first constraint set and second constraint set;Re-execute the step of obtaining target image until the number of vanishing points in the first constraint set and the second constraint set reaches the preset threshold, obtain target first constraint set and target second constraint set;According to the target first constraint set and the target second constraint set, the rotation extrinsic parameter of the target vehicle is optimized.The camera and vehicle extrinsic parameter online calibration optimization method provided by the application can make the sensor data obtained by the user when using the target vehicle more accurate, and drive more safely.

The invention relates to a novel digital intelligent ship type irrigation device which comprises an upper ship body and a lower ship body which are connected with each other. The upper ship body is provided with a photo-thermal solar power generation panel, a 360-degree omnibearing camera, a second positive and negative rotation motor and a spray head; the lower ship body comprises a ship middle part and a ship bottom part; the ship middle part is provided with a sensor, a centrifugal pump, a low-rotating-speed angle-controllable gear motor, a first positive and negative rotation motor and a battery; the ship bottom part is provided with a rotatable propelling module, and the rotatable propelling module comprises a ship bottom rotating bearing, a ship bottom propeller, a ship bottom lower connecting bearing rod and a ship bottom upper connecting bearing rod; accurate control of the propelling direction is realized; the spraying angle of the sprayer and the rotating speed of the centrifugal pump are fed back through a sensor to achieve linkage adjustment. Compared with the prior art, dual-module power supply is achieved, and the resource utilization rate is increased; intelligent and accurate control is realized, and manpower is saved; and the irrigation area can be dynamically adjusted.

The present invention discloses a mobile robot relocalization method based on global descriptors. In the relocalization stage, through preprocessing the data of the omnidirectional camera sensor system on the mobile robot, a local dense point cloud containing semantic information is obtained; a global descriptor is created for the above local dense point cloud, and it is matched and extended with a dense point cloud sub-map set, and the sub-map with the highest similarity score among them is used as the target dense point cloud; through point cloud registration of the local dense point cloud and the target dense point cloud, the relocalization result of the mobile robot is obtained. The invention fully integrates the semantic information and metric information provided by the omnidirectional camera sensor system, and achieves accurate and robust relocalization effects for mobile robots in complex environments with large-scale ranges, across indoor and outdoor scenarios, and in scenarios with similar geometric features.

Methods and processors for rendering a 3D Gaussian are disclosed. A processor is configured to: acquiring the 3D Gaussian to be rendered, and a camera model representing an omnidirectional camera with an optical axis; updating a color of the 3D Gaussian using spherical harmonics; updating a position of the 3D Gaussian by moving the 3D Gaussian towards the optical axis; updating a scale of the 3D Gaussian by compressing the 3D Gaussian in at least one of a tangential direction and a polar direction relative to the optical axis; and rendering an updated 3D Gaussian onto a 2D plane using a 3DGS model.

A virtual object display device (1) comprises a video generation unit (14) that: uses, as a virtual object, a physical body object representing a physical body and a spherical object representing an omnidirectional camera; is connected via a network to a server (5) that stores an omnidirectional video obtained by capturing the real space by the omnidirectional camera and the position information of the virtual object; and on the basis of the position information of the virtual object, generates and displays as a display video an MR video obtained by mixing the omnidirectional video with a video obtained by capturing the virtual object by a virtual camera disposed at the center of the omnidirectional camera, or a virtual video obtained by capturing the virtual object by the virtual camera disposed at an arbitrary position, and a video switching unit (130) that switches the display video between the MR video and the virtual video on the basis of an operation instruction.

A method includes the pairing of a head-mounted virtual reality device with an omnidirectional camera, the display of a converted media field to a user associated with media content corresponding to an external environment relative to a vehicle, the generation of one or more augmented-reality images, and the superimposition of the one or more augmented-reality images onto the displayed converted media file.

A method includes the pairing of a head-mounted virtual reality device with an omnidirectional camera, the display of a converted media field to a user associated with media content corresponding to an external environment relative to a vehicle, the generation of one or more augmented-reality images, and the superimposition of the one or more augmented-reality images onto the displayed converted media file.