49 results about "Mobile vision" patented technology

The invention provides a crop population three-dimensional reconstruction and organ phenotypic character analysis method, and relates to the field of three-dimensional reconstruction, and the method comprises the steps: collecting cotton envelope image data, carrying out the data preprocessing, and carrying out the motion recovery processing and Gaussian sputtering of the cotton envelope image data, the method comprises the following steps of: preprocessing a group cotton point cloud, removing a ground point cloud, obtaining a complete cotton group point cloud model, segmenting a cotton individual point cloud model and a corresponding organ point cloud model from the complete cotton group point cloud model, automatically analyzing cotton key phenotypic characters, and performing correlation analysis. According to the method, a cotton original habitat three-dimensional point cloud model is constructed based on mobile visual equipment, and key characters such as cotyledon node height, plant height and leaf area can be automatically analyzed, so that the problems that phenotypic character extraction in the prior art is high in cost, low in flux, easy to make mistakes, inconsistent in standard, generally destructive, labor-consuming, time-consuming and the like are solved.

The invention discloses a battery replacing station and a control method of the battery replacing station. The battery replacing station comprises a controller, a driving device, a grabbing device, a visual device and at least two battery bins, and the controller is in communication connection with the driving device, the grabbing device and the visual device; the visual device and the grabbing device are both arranged on the driving device, the driving device drives the visual device and the grabbing device to move in the working space of the battery replacing station, the visual device is used for obtaining image information of a target object to recognize the spatial pose of the target object, and the battery bin is used for storing batteries; the controller is configured to execute the following operations: obtaining a first pose of the power-deficient battery, a second pose of the idle battery compartment and a third pose of the full-power battery compartment through the visual device; and based on the first pose of the power-deficient battery, the second pose of the idle battery bin and the third pose of the full-power battery bin, the driving device and the grabbing device are controlled to execute grabbing or releasing operation on the target battery. The battery replacement efficiency can be improved.

The invention provides a high-precision probe assembly system and method, and the system is characterized in that a probe assembly suspension clamp is fixed on an installation platform, and the probe assembly suspension clamp is used for suspending a probe and controlling the probe to press down and adhere to a suspension structure; the three-axis displacement table is fixed on the mounting platform, the three-axis displacement table is located below the probe assembly suspension clamp, one surface, facing the probe assembly suspension clamp, of the three-axis displacement table is used for fixing a suspension structure, and the three-axis displacement table can move in the horizontal or vertical direction; the visual acquisition module is fixed on the mounting platform and is used for acquiring a relative position image; the three-axis displacement table is used for moving by referring to the relative position image. According to the high-precision probe assembly system and the assembly method, the alignment process of the probe and the central circular surface of the suspension structure is fully automatically realized, the alignment precision is high, and the speed is high; a specially-made probe assembly suspension clamp is matched, the probe is controlled to be vertically pressed downwards, and finally accurate bonding assembly of the probe and the suspension structure is achieved.

The utility model discloses a mobile vision AI safety early warning device which comprises a vision early warning box body, a horizontal adjusting base is arranged at the bottom of the outer side of the vision early warning box body, electric telescopic rods are arranged at the two ends of the inner side of a square frame at the end of the horizontal adjusting base, and a monitoring camera moves downwards through cooperation of a servo motor and a threaded vertical rod. The monitoring camera enters the cleaning groove and rotates in the cleaning groove through a micro motor, surface dust cleaning is conducted through a cleaning soft rubber brush arranged in the cleaning groove, the cleaned dust is adsorbed through a dust remover by making the cleaning groove generate suction force through a dust removal pipe fitting, and the adsorbed dust is collected in a dust collection box. And therefore, the problems that the environment is affected and the view of the monitoring camera is blurred due to the fact that dust is diffused to the periphery during cleaning are avoided, and the convenience of cleaning the monitoring camera is further improved.

The application provides a walking robot on a steel mesh, and relates to the technical field of building robot equipment. The walking robot on the steel mesh comprises a chassis, a moving vehicle, a first space, a rack, a first transmission unit, a first driving unit, a second driving unit, a visual recognition unit and a motor. The chassis and the moving vehicle are located on the steel mesh, the motor provides driving force for the visual recognition unit, the second driving unit or the first transmission unit, the second driving unit drives the moving vehicle to move on the chassis in the process, and the moving vehicle drives the first transmission unit to move on the rack in the first space, and the visual recognition unit transmits the obtained obstacle data information and movement data information to the control system through the control module. Therefore, when working on the steel mesh, the robot is not affected by the diameter and spacing of the steel bars, the steel bar binding and the steel bar welding and other problems in the moving process, and the movement of the robot in the working process is not blocked.

The application belongs to the technical field of mobile visual inspection planning, and specifically discloses a mobile visual inspection station-point joint clustering and sequence planning method. The method is a genetic algorithm based on an alternating evolution strategy, and comprises steps of data encoding / decoding, evolution strategy updating, individual selection, individual crossover, individual mutation, population updating and the like. The length of a chromosome is equal to the number of points participating in sequencing. The integer part of a gene represents a point sequence number, and the decimal part represents a station sequence number to which the point belongs. Point sequences belonging to the same station follow the order of appearance in the chromosome; station sequences follow the order of the first appearance position. The application can integrally realize station-point clustering and sequence planning, and improve visual inspection efficiency; the proposed evolution strategy also significantly reduces the running time of the planning program. The application is suitable for large and complex component mobile visual inspection and other similar scenarios.

The invention relates to a display system (100) for a vehicle (10), comprising: - a driver monitoring module (110) which is configured to detect the viewing behaviour of a driver of the vehicle (10); - a display module (120) which is configured to display a moving visual content (BVI); and - a control module (130) which is configured to actuate the display module (120) in such a way that a playback speed of the moving visual content (BVI) is adjusted, in particular reduced, on the basis of the viewing behaviour.

This invention relates to a fast selection interaction method, apparatus, and electronic device for an electronic paper display, comprising: rendering and displaying an interactive interface containing multiple candidate options on an electronic paper display unit, wherein the interactive interface does not display a movable visual focus used to indicate the current candidate option; establishing a corresponding mapping relationship between each option in the multiple candidate options and each specific physical direction button, the mapping relationship being determined based on the spatial arrangement of each option in the interactive interface; responding to a detected trigger operation on a target physical direction button in a group of physical direction buttons, identifying the trigger operation as a selection operation on the option mapped to the target physical direction button according to the mapping relationship, and executing the subsequent logic corresponding to that option; during the process of the trigger operation being responded to and the subsequent logic being executed, interface refresh triggered by moving the movable visual focus is not performed. This invention can avoid frequent screen refreshes and operation delays caused by cursor movement.

The invention discloses digital twin automatic wiring equipment based on fusion of a large model algorithm and an AI vision algorithm, and belongs to the technical field of automatic wiring equipment. The equipment comprises an execution module, an AI visual acquisition module, a digital twinborn data acquisition module, a communication module and a digital twinborn management and control platform integrated with a dual-algorithm module, the execution module realizes wire clamping and three-dimensional movement, the AI visual acquisition module completes accurate positioning and process tracking, the digital twin data acquisition module acquires equipment operation state data in real time, the communication module realizes data interaction and instruction transmission, and the digital twin management and control platform constructs an equipment virtual mapping model. And parameter self-optimization, fault prediction and virtual-real cooperative regulation and control are realized. The problems that traditional equipment is inaccurate in positioning, poor in adaptability, lagged in regulation and control and the like are solved, the wiring quality and the operation efficiency are remarkably improved, the operation and maintenance cost is reduced, and the method is suitable for wiring scenes of assembly of various electrical equipment.

A mobile vision transformer network for use on mobile devices, such as smart eyewear devices and other augmented reality (AR) and virtual reality (VR) devices. The mobile vision transformer network considers factors including number of parameters, latency, and model performance, as they reflect disk storage, mobile frames per second (FPS), and application quality, respectively. The mobile vision transformer network processes images, e.g., for image classification, segmentation, and detection. The mobile vision transformer network has a fine-grained architecture including a search algorithm performing latency-driven slimming that jointly improves model size and speed.

This invention belongs to the field of building deviation detection technology, and particularly relates to a method for detecting building settlement and displacement deviation based on visual processing. The method first selects natural and artificial landmarks with long-term stability, and after AI visual stability classification, constructs a global benchmark landmark library by combining it with absolute geodetic coordinates. It then acquires multi-view sequential images of the target building and landmarks through a mobile visual acquisition platform, constructs a unified global absolute coordinate system using the benchmark landmarks as global constraints, and generates a high-precision 3D digital model of the target building with absolute 3D coordinates. The model is repeatedly acquired and modeled at a preset cycle, and multi-cycle model registration is completed using stable landmarks as anchor points. Differential calculations are then used to solve the building's full-dimensional deformation deviation data, completing compliance judgment and graded early warning. This invention solves the pain points of traditional detection benchmarks being prone to failure and incomplete detection dimensions, significantly improving the accuracy and timeliness of building deformation detection and early warning.

The invention belongs to the technical field of inspection robots, and particularly relates to a wheel type inspection robot with an anti-collision function, which comprises a loading body, a driving protection shell and a visual sensor. According to the invention, the robot needs to complete preparatory work before operation, including checking the electric quantity of a battery to ensure that the electric quantity is sufficient, verifying that the functions of a visual sensor and a camera are normal, confirming that a navigation path has no obstacle, and confirming that a protection plate is in an embedded and buckled fixed state, and a supporting plate 8 is locked and folded by a buckle plate 13. The loading body moves along a preset route through the driving wheel and the driven wheel, and the visual sensor collects environmental data in real time and uploads the environmental data to the management platform; if one side of the loading body is impacted, the impact force is transmitted to the sleeve through the impact side protection plate, the trigger rod on the other side is pushed after the second spring is compressed, the opposite protection plate is separated from the constraint of the buckle plate, at the moment, the supporting plate rapidly slides downwards and stretches out by means of the elastic force of the first spring, and flexible supporting is formed after the inserting plate and the inserting rod reach the maximum stroke. And meanwhile, the trigger rod automatically resets and is centered under the springback effect of the second spring.

The application discloses a kind of wet process equipment mobile vision protection devices, comprising: protective cover, protective plate, sealing protection ring and air pipe assembly;Protective plate is fixedly connected at one end of protective cover, and is provided with mounting through hole;Sealing protection ring includes: first ring body and second ring body;First ring body and second ring body are interconnected;Sealing protection ring is sealingly connected with mounting through hole;Multiple overflow notches, multiple overflow flow channels, multiple flow guide notches are opened in first ring body, inner wall is sealingly connected with lens mounting portion;Annular positive pressure gas static pressure distribution groove is opened in the inner wall of the connecting portion of first ring body and second ring body;Second ring body is matched with the gap of lens;Visual camera is fixedly arranged in protective cover.The gas is distributed by two-stage static pressure of protective cover and sealing protection ring, and forms low-speed annular laminar flow air curtain along the cylindrical outer wall of lens, protects the field of view near camera lens, and the flow is controllable, low-flow energy-saving, improves the online operation rate of mobile vision.

The invention discloses an intelligent industrial high-precision mobile visual inspection robot, and belongs to the technical field of inspection robots, the intelligent industrial high-precision mobile visual inspection robot comprises a fixed seat, an electric sliding rail is fixedly mounted on the periphery of the fixed seat, an annular sliding block is mounted in the electric sliding rail, a mechanical arm is mounted on the annular sliding block, and a mounting seat is fixedly mounted on the mechanical arm; a driving motor for driving is embedded in the mounting seat, a rotating block is arranged at the output end of the driving motor, wiping cotton is mounted on the rotating block, an image acquisition module is arranged on the lower side of the mounting seat, a data processing module is mounted on the fixed seat, a control module is further mounted on the fixed seat, and the control module is used for controlling the mechanical arm and the driving motor. According to the device, the wiping cotton and the image acquisition module are matched with the data processing module, so that whether stains exist on workpieces or not can be judged, whether the stains are stuck in the carrying process or not can be judged, whether the stains are embedded in the workpieces or not during machining or not can be judged, and the workpieces can be effectively screened.

The invention relates to a rapid option interaction method and device for electronic paper display equipment and electronic equipment, and the method comprises the steps: rendering and displaying an interaction interface containing a plurality of options to be selected on an electronic paper display unit, and not displaying a movable visual focus used for indicating a current candidate option on the interaction interface; establishing a corresponding mapping relationship between each option in the plurality of options to be selected and each key in the specific physical direction, wherein the mapping relationship is determined based on the spatial arrangement layout of each option in the interactive interface; in response to a detected trigger operation on a target physical direction key in the physical direction key group, identifying the trigger operation as a selection operation on an option mapped by the target physical direction key according to the mapping relationship, and executing subsequent logic corresponding to the option; in the process that the trigger operation is responded and the subsequent logic is executed, interface refreshing triggered by moving the movable visual focus is not executed. According to the invention, frequent screen refreshing and operation delay caused by cursor movement can be avoided.

The application discloses a shell separation process of an electric energy meter, and relates to the technical field of electric energy meter disassembly, which comprises the following steps: AGV forklift transfers a tray containing an electric energy meter to a disassembling machine, a conveying line conveys the tray, a feeding robot moves the electric energy meter to a rotary table type disassembling station, removes the connecting screw of an upper cover and a lower cover (three for the upper cover and one for the lower cover), clamps the upper cover to a sub-piece disassembling line, separates the tail of the meter from the lower cover and transfers the lower cover, removes the screw of a circuit board, separates the upper circuit board, visually positions the screw of the lower circuit board, clamps the lower circuit board to the sub-piece disassembling line, and stacks and recycles empty trays, and under the assistance of a 3D visual system and a movable visual positioning camera, the automatic disassembly of the electric energy meter can be realized, the whole disassembly process is closed-loop operation relying on automatic equipment, large-scale disassembly operation can be stably completed, the disassembly efficiency is improved, and the problem of low operation standardization degree caused by excessive dependence on manual experience in the traditional operation mode is solved.

The utility model relates to a kind of 3D printing equipment with visual defect detection, belong to industrial visual detection field.The equipment includes base, 3D printing part and two visual detection parts, linear slide is horizontally arranged in the middle part of base and build plate, and build plate is driven to move left and right by first motor;3D printing part adopts linear guide rail, ball screw and synchronous belt transmission structure, and three-dimensional accurate movement of printing head is realized by cooperation motor drive;Visual detection part is equipped with vertical frame, movable camera support and visual detection camera connected by universal joint, and two visual detection cameras are cross arranged left and right.The utility model realizes the integration of printing and visual defect detection, and can complete detection without taking down part, with compact structure, high degree of automation, the characteristics of comprehensive accurate detection, effectively improve 3D printing part production efficiency and quality control level, applicable to multi-field 3D printing production.

This invention relates to a method for mobile vision detection of construction workers' protective gear. The method uses YOLO11m as the basic architecture to build a detection model. The backbone of the detection model includes an SPFF module, and a reflection-sensitive feature channel is introduced at the end of the backbone. The head of the detection model incorporates a conditional attention mechanism, which activates local attention weights in the region below the safety helmet in the image when a safety helmet is detected. The SPFF module can fuse deep semantics and shallow detail features for small targets at 1K+ resolution through multi-scale dilated convolution and channel attention, overcoming the bottleneck of traditional FPN in losing features of small targets and significantly enhancing the feature representation capability of small targets.

The utility model relates to a kind of dual-function detection equipment, comprising: bearing mechanism, including bearing seat and translation drive component, the bearing seat is used to place product to be detected, the translation drive component is used to drive the bearing seat moves between feeding position and detection position;Visual inspection mechanism, erect in the above of the detection position;And line sequence detection mechanism, be located at the detection position, including pressing plate and lifting drive component, the lifting drive component is used to drive the pressing plate vertical movement, to make the pressing plate compact the product and with the product electric connection, the pressing plate is equipped with several through holes, the through hole exposes the product for the visual inspection mechanism detection.The utility model integrates line sequence detection and appearance detection in one, equipment space layout can be optimized, effectively promote detection efficiency.

A method for stereoscopic telestration in a 3D display system, and system therefore includes the steps of, and hardware for, aligning a telestration base plane to a reference object visible in a stereoscopic video stream; fusing a telestration overlay at the telestration base plane to ensure accurate depth alignment; and repositioning the telestration overlay to a moved telestration plane in front of or behind the telestration base plane by adjusting left and right eye overlays symmetrically. The repositioning may be performed without losing positional accuracy relative to the stereoscopic image stream.

The utility model relates to the technical field of detection devices, and particularly discloses a double-track double-material conveying type detection device which comprises an equipment main body, the equipment main body comprises a bearing frame, a visual detection device, a double-track conveying device arranged on the bearing frame and a driving device, the driving device is used for driving the visual detection device, and the double-track conveying device comprises a first conveying track and a second conveying track; each of the first conveying track and the second conveying track comprises a rotationally arranged conveying belt and a first driving module for driving the conveying belt, the conveying belts are used for bearing an external material carrying disc and driving the external material carrying disc to move, and the visual detection device is used for photographing or shooting materials in the external material carrying disc to distinguish qualified materials from unqualified materials; the visual inspection device photographs or photographs the materials borne and conveyed by the first conveying track and the second conveying track so as to distinguish qualified materials and unqualified materials, the advantage of detecting different materials at the same time is achieved, and the production cost can be effectively saved.

The invention provides a lithium ion battery state estimation method based on an advanced mobile vision field. The method comprises the following steps: establishing an equivalent circuit model: establishing a lithium ion battery equivalent circuit model containing a hysteresis effect and temperature related parameters; constructing a nonlinear state space model based on the equivalent circuit model, wherein the state space model comprises a state equation for describing dynamic behaviors of the system and an observation equation for reflecting voltage response; an event trigger linearization mechanism (ETR) is introduced into a mobile vision field estimation (MHE) optimization framework; dynamically adjusting the weight of the cost function by combining a Gaussian-Newton iteration (GN) optimization method; and a high-precision SOC estimation result is output by using real-time measurement data through complete modeling, prediction and optimization processes. On the basis that the advantages of an MHE optimization structure are reserved, battery physical nonlinear factors are further fused, and the SOC estimation method with temperature adaptability solves the problem of estimation deviation caused by the fact that a traditional MHE simplified model ignores physical characteristics.

The utility model provides a handheld mobile visual inspection device and an inspection system. The handheld mobile visual inspection device comprises an assembling mechanism; the positioning mechanism comprises a distance sensor and a laser generator, and the distance sensor and the laser generator are arranged towards the to-be-detected element; and the recognition mechanism comprises a recognition camera, the recognition camera is connected to the assembly mechanism and faces the to-be-detected element, and the recognition camera is in signal connection with the distance sensor and the laser generator. In the scanning detection process, the positioning mechanism positions the distance of the to-be-detected element through the distance sensor, and the laser generator positions the detection angle and direction of the to-be-detected element, so that the recognition mechanism can only be recognized in a specific position, and the detection precision of the handheld mobile visual detection device is greatly improved; in addition, compared with a conventional detection technology at the present stage, the method has the remarkable advantages of being high in compatibility, flexible to use, convenient to operate, small in detection error and the like.

The invention relates to the technical field of restaurant robots, and discloses an integrated restaurant robot which is characterized in that a storage cleaning cavity and a mounting cavity are formed in a main body, a storage assembly, a cleaning assembly and a heating assembly are arranged in the storage cleaning cavity, a control assembly is arranged in the mounting cavity, and the control assembly is connected with the cleaning assembly and the heating assembly; the mechanical arm is mounted on one side of the main body, is connected with the control assembly and is used for grabbing, placing and transferring tableware; the moving mechanism is connected to the lower part of the main body, is connected with the control assembly and is used for realizing overall movement; the visual monitoring camera is arranged on the upper portion of the main body, connected with the control assembly and used for recognizing the position of tableware and cooperating with the mechanical arm to complete actions, the limitation that a traditional robot is single in function is broken through, and full-process automatic operation from meal delivery and meal collection to cleaning and drying is achieved.

A method includes, while a vehicle is restricted to a parked condition, presenting a visual tracking application on a display of the vehicle, the visual tracking application including a moving-target user interface (UI) element and a pursuit UI element, detecting, using a steer-by-wire (SbW) steering system of the vehicle, movements of a steering wheel by an operator of the vehicle, moving, based on the detected movements, the pursuit UI element in the visual tracking application, and determining, at a plurality of times, a corresponding tracking error between a location of the moving-target UI element in the visual tracking application and a location of the pursuit UI element in the visual tracking application. The method also includes processing the corresponding tracking errors to estimate a visual-motor skill of the operator.

The invention discloses a movable visual acquisition device for detection of a direct current converter valve assembly, and relates to the field of power system detection equipment, and the movable visual acquisition device comprises a mounting base, a first translation screw rod module, a rotating seat, a cantilever, a cantilever angle limiting mechanism, a second translation screw rod module and a visual acquisition unit, the installation base is used for being detachably installed on a frame of a converter valve assembly and comprises two fixing bases, the two ends of the first translation lead screw module are fixedly connected to the two fixing bases respectively, the rotating base is arranged on a moving sliding table of the first translation lead screw module, one end of the cantilever is rotationally connected to the rotating base, and the other end of the cantilever is rotationally connected to the rotating base. The two cantilever angle limiting mechanisms are fixedly arranged on a movable sliding table of the first translation lead screw module in a 90-degree mode and used for limiting and locking the positions of the cantilever in the storage state and the working state, and the second translation lead screw module is arranged on the cantilever in the lengthwise direction of the cantilever. And the visual acquisition unit is mounted on a movable sliding table of the second translation screw rod module. According to the invention, the converter valve detection speed and efficiency are greatly improved.

The invention provides an intelligent inventory checking system and method based on mobile vision, and the system comprises a cabinet body, a cabinet door, a guide rail mechanism which is disposed at a cabinet opening, and the laying position of the guide rail mechanism is matched with the cabinet opening. The charging mechanism is arranged at the cabinet opening, and the charging mechanism is provided with a charging position; the checking mechanism comprises a camera module, a driving module, a wireless communication module, a battery module and a control module; the checking mechanism is in sliding connection with the guide rail mechanism through a driving module and has a standby state and a checking state relative to the cabinet body; when the checking mechanism is in a checking state, the checking mechanism moves for a circle along the guide rail mechanism, and the camera module shoots and collects commodities in the cabinet body and obtains multi-angle image information of the commodities; the control module is electrically connected with the camera module and is used for receiving and processing the multi-angle image information, obtaining inventory list information and processing the inventory list information to generate order information; and when the checking mechanism is in the standby state, the checking mechanism is in the charging position, and the charging mechanism charges the checking mechanism.

The present invention relates to a high-efficiency and long-service-life dimmable organic light-emitting device and a use thereof. The light-emitting device comprises: a light-emitting layer. In the light-emitting layer, two or more host materials are mixed together and a luminescent material is doped in the layer; the luminescent material is a platinum complex, and the luminescent material emits light in a monomer emission state and at least one aggregation emission state; the LUMO energy level of at least one host material is lower than the LUMO energy level of the luminescent material or the difference between the LUMO energy level of at least one host material and the LUMO energy level of the luminescent material is less than 0.2 eV, or the host material in the light-emitting layer employs a combination of SF-BCZ and SF-TRZ. The light-emitting layer or the light-emitting device can be applied to fixed visual display units, mobile visual display units, lighting units, keyboards, clothing, decorative items, clothing accessories, wearable devices, medical monitoring devices, wallpapers, tablet computers, laptops, advertising panels, panel display units, household appliances, and office supplies.

A method for stereoscopic telestration in a 3D display system, and system therefore includes the steps of, and hardware for, aligning a telestration base plane to a reference object visible in a stereoscopic video stream; fusing a telestration overlay at the telestration base plane to ensure accurate depth alignment; and repositioning the telestration overlay to a moved telestration plane in front of or behind the telestration base plane by adjusting left and right eye overlays symmetrically. The repositioning may be performed without losing positional accuracy relative to the stereoscopic image stream.

A mobile vision transformer network for use on mobile devices, such as smart eyewear devices and other augmented reality (AR) and virtual reality (VR) devices. The mobile vision transformer network considers factors including number of parameters, latency, and model performance, as they reflect disk storage, mobile frames per second (FPS), and application quality, respectively. The mobile vision transformer network processes images, e.g., for image classification, segmentation, and detection. The mobile vision transformer network has a fine-grained architecture including a search algorithm performing latency-driven slimming that jointly improves model size and speed.