38 results about "Linear object" patented technology

The invention relates to the technical field of data processing, in particular to a robot control method, device and system based on a diffusion model and a medium, and the method comprises the steps: collecting the position information of the tail end of an index finger and the state information of a flexible linear object in a human hand teaching process, and constructing a training sample; training the training sample by using a diffusion model, and learning mapping from a high noise state to a clear target state through a noise prediction network in the training diffusion model to obtain a trained diffusion model; the training process of the diffusion model comprises forward noise addition and reverse noise reduction; the state of the flexible linear object and the tail end position of the mechanical arm are collected, a current observation environment is constructed, the current observation environment is input into the trained diffusion model, a target action sequence is generated, and the mechanical arm is controlled to execute control actions based on the target action sequence; according to the invention, the accuracy and real-time performance of the robot to control the flexible linear object can be improved.

The invention relates to the technical field of industrial detection, in particular to an improved YOLOv5-based safety helmet standard wearing detection method, which comprises the following steps of S1, establishing a safety helmet wearing data set with labels; s2, modifying an original feature extraction neural network CSPDarknet to obtain a pre-trained neural network, and performing secondary training on the pre-trained neural network according to the training set to obtain a trained neural network; s3, verifying the neural network training effect according to the test set to obtain a final neural network; and S4, obtaining a safety helmet wearing detection result according to the working video of the safety production area and the final neural network. According to the method, the adaptability of the non-rigid linear object is enhanced through the snakelike convolutional network; the features of the semantic layer are made to pay attention to global information through a transform layer; the optimization feature enhancement module is an AFPN to realize the characterization capability of the overall enhanced network; for a standard wearing scene of the safety helmet, soft-mosaic is proposed in data enhancement to enrich sample diversity, so that the model is more robust and has generalization.

The invention discloses a method for calculating image force feedback by physically modeling a guide wire based on a discrete elastic rod, which comprises the following steps of: modeling the guide wire into a non-telescopic deformable linear object, representing the center line of the guide wire by adopting an arc length parameterized three-dimensional curve, and describing bending and torsion characteristics by combining an orthogonal material frame; a curvature vector and a torsion angle are calculated through a discretized particle model to simulate the internal state of the guide wire, and the motion of the guide wire is predicted based on a semi-implicit integration method through a Lagrange motion equation; then, coarse detection is carried out in combination with the preprocessed mask image of the blood vessel wall, whether the guide wire makes contact with the blood vessel wall or not is judged, and due to the existence of a preprocessing structure, the detection is faster than a common bounding box which is divided through a tree structure; then, whether the tip of the guide wire penetrates through the blood vessel wall or not is judged through the blood vessel center line and kd-tree, a force feedback calculation model is constructed by combining a spring damping system according to deformation caused to the blood vessel wall, and corresponding feedback force after collision is calculated.

The invention discloses a ground fracture detection method based on remote sensing image visual analysis, which comprises the following steps of: sending a remote sensing image to a trained ground fracture dynamic monitoring model, outputting a fracture path by the ground fracture dynamic monitoring model, outputting a vector element graph layer containing a detected fracture position, and identifying a section of path identified as a fracture corresponding to each element, displaying through a GIS platform; the construction and training process of the ground fracture dynamic monitoring model comprises the following steps: S1, constructing the ground fracture dynamic monitoring model, and manually marking the accuracy of a detection result of a DOM image sample; s2, improving the recognition precision of the fine-line object target in the model structure; s3, optimizing the extraction efficiency of the interference features; s4, obtaining landform parameters of a to-be-monitored area, and introducing a high-speed interpretation image sample; and S5, obtaining a trained ground fracture dynamic monitoring model. The invention belongs to the technical field of geological image processing, and significantly improves the efficiency and accuracy of ground fracture recognition.

The invention discloses a nonlinear PID control method based on intelligent element driving and a fault-tolerant control system thereof, and belongs to the technical field of intelligent control. According to the method, three behavior mapping mechanisms of punishment, memory and prediction are introduced, a traditional PID controller structure is expanded, a compound error signal is constructed, and a dual-channel control strategy of instinct response and learning response is designed. The controller provided by the invention has adaptive gain adjustment capability, a dynamic adjustment law is constructed in combination with a Lyapunov stability theory, and robust control without model information is realized. Meanwhile, a fault-tolerant mechanism is provided for correcting control signal input by considering the situation of performance degradation of an actual actuator, so that the stability and tracking precision of a control system under non-ideal conditions are improved. The control method is suitable for complex nonlinear object control scenes such as underwater robots, mobile robots and parallel Delta robots.

The application provides a remote sensing image road edge detection method and system, and the method comprises the following steps: extracting a predicted candidate vertex set Q from a road image; selecting a candidate vertex from the predicted candidate vertex set Q as a predicted starting vertex v i *, generating a local feature input agent model thereof, outputting a next predicted vertex v i+1 *, and updating a generated road edge graph g i+1 * based on the next predicted vertex v t *, continuing prediction until the road edge graph g t stops growing; and traversing each candidate vertex in the predicted candidate vertex set Q, performing prediction based on the agent model to obtain a road edge detection result. The method mainly adopts an end-to-end iterative graph generation method, can directly obtain linear object information, is less affected by segmentation performance, can solve the influence caused by partial occlusion, and can obtain vector data of the road edge without complicated post-processing operations.

This invention relates to the field of data processing technology, specifically to a robot manipulation method, device, system, and medium based on a diffusion model. The method includes: collecting the position information of the index finger tip and the state information of a flexible linear object during a human hand teaching process to construct training samples; training the training samples using a diffusion model, and learning the mapping from a high-noise state to a clear target state through a noise prediction network in the training diffusion model to obtain a trained diffusion model; the training process of the diffusion model includes forward noise addition and backward noise reduction; collecting the state of the flexible linear object and the position of the robotic arm end effector to construct the current observation environment, inputting the current observation environment into the trained diffusion model to generate a target action sequence, and controlling the robotic arm to perform manipulation actions based on the target action sequence; this invention can improve the accuracy and real-time performance of robot manipulation of flexible linear objects.

1. The name of this design product: Linear object retracting and releasing device (gripper). 2. Purpose of this design product: used for reeling in and unreeling in wires. 3. The key point of the design of this product lies in its shape. 4. The picture or photo that best illustrates the design points: three-dimensional picture.

The present invention belongs to the field of image processing and provides a method for fusing satellite and low-altitude unmanned aerial vehicle remote sensing data. The method comprises the following steps: extracting linear object contour information from a satellite remote sensing image of a target area and dividing the linear object contour into a plurality of contour segments; extracting real-time linear object contour information from the unmanned aerial vehicle remote sensing image of the target area and dividing the real-time linear object contour into a plurality of contour segments with the same structural form as in the satellite remote sensing image; performing corresponding matching on the linear object contour segments in the satellite remote sensing image and the unmanned aerial vehicle remote sensing image to construct a stable contour segment pair; performing geometric registration processing on the satellite remote sensing image and the stable area in the unmanned aerial vehicle remote sensing image; adjusting the observation angle of the unmanned aerial vehicle image in the changed area based on the relative spatial relationship between the unmanned aerial vehicle flight trajectory and the linear object, and fusing the image after the angle adjustment with the satellite remote sensing image.

In a plan creation support method disclosed herein, an arrangement plan including arrangement paths connecting respective starting points and end points of a plurality of linear objects in a predetermined area is created by investigating the arrangement paths on the basis of linear object information, including information on the respective starting points and end points of the plurality of linear objects, a plurality of requests, which function as constraining conditions when a business operator arranges the plurality of linear objects in the predetermined area, and area information defining the predetermined area. Moreover, a basic request achievement degree indicating the degree to which the arrangement plan satisfies the plurality of requests is calculated, and when individual requests are excluded from the plurality of requests, a second request achievement degree indicating the degree to which the arrangement plan satisfies the plurality of requests excluding the individual requests is calculated for each request. A request influence degree, which is an evaluation index of how much the second request achievement degree is improved relative to the request achievement degree, is calculated for each request on the basis of the second request achievement degree for the request and the basic request achievement degree.

The utility model relates to the technical field of wire coils, in particular to a wire winding and unwinding device which comprises a mandrel and two discs, one end of each disc is fixedly connected with a clamping ring, and two ends of the mandrel are respectively clamped on the clamping rings of the two discs. A rotating hole is formed in the center of the disc, and handle grooves are formed in the left side and the right side, close to the rotating hole, of the disc. The utility model has the advantages that by arranging the handle groove, when the device needs to be carried, a single hand can be inserted into the handle groove to pick up the device for carrying, and the fillets in the handle groove can prevent fingers from being scratched, so that the aim of protecting the fingers is fulfilled; through the arrangement of the wire hanging plate, the inclined wire hanging plate and the wire hole can more conveniently penetrate through a thread, and the purpose of conveniently fixing the end part of the thread is achieved; and by arranging the positioning columns and the positioning holes, quick positioning is facilitated when the device is stacked.

The present application belongs to the field of visual tracking technology, and discloses a kind of processing existing partial occlusion flexible linear object visual tracking method, based on RGB-D camera obtains color image and depth image data, from first frame image, by mask segmentation, skeletonization processing operation generates initial state point, in each subsequent frame, object point cloud is generated by mask in conjunction with depth image;Then, the density of its surrounding current point cloud is counted, and the state point is marked as "visible" or "occlusion";By establishing the analytical model based on the spatiotemporal continuity of flexible linear object, the closed form solution of the occlusion position is derived, and the iterative optimization process of the occlusion position is eliminated, thereby reducing the computational complexity and achieving efficient real-time tracking;Meanwhile, using the inherent spatiotemporal geometric evolution law of flexible linear object, a position estimation model that can be directly analytically solved is constructed, avoiding the stability risk brought by iterative optimization, and significantly improving the reliability and tracking stability of the occlusion position estimation.

This linear object detection device comprises: a straight line detection unit that detects, with respect to a three-dimensional point cloud representing three-dimensional coordinates at points on the surface of a structure, straight lines on the basis of the three-dimensional coordinates, divides the three-dimensional point cloud related to one of the detected straight lines according to the distance between three-dimensional points related to the one detected straight line, and detects straight lines again on the basis of the three-dimensional coordinates for each divided three-dimensional point cloud; and a classification unit that groups the straight lines on the basis of the orientations of the plurality of straight lines detected and angles between the straight lines and the ground, and classifies straight lines among the straight lines included in the same group as straight lines corresponding to the same structure if angles between the straight lines and line segments connecting the straight lines are within a threshold value.

The present application provides a method, apparatus, device, and storage medium for processing altitude data of a high-precision map. Among them, the method includes: obtaining the identification ID of a target object in the high-precision map and the OID value carried by the identification ID, where the target object includes road traffic facilities; converting the target object into a linear object and obtaining the original altitude of each shape point in the linear object; determining the fuzzy altitude of each shape point respectively based on the OID value of the target object and the original altitude of each shape point, and generating the fuzzy altitude data of the target object based on the fuzzy altitude of each shape point. Through the above method, the consistency of the deviation of each fuzziness of the same target object in the high-precision map is effectively improved, achieving the purpose of enhancing the usage effect of the high-precision map by users.

The invention belongs to the technical field of visual tracking, and discloses a visual tracking method for processing a partially-shielded flexible linear object, which comprises the following steps of: acquiring color image and depth image data based on an RGB-D camera, generating an initial state point from a first frame of image through mask segmentation and skeletonization processing operation, and in each subsequent frame, generating an initial state point from a second frame of image through mask segmentation and skeletonization processing operation; generating an object point cloud by combining the mask with the depth image; then, counting the density of current point clouds around the state point, and marking the state point as visible or shielded; an analytical model based on space-time continuity of a flexible linear object is established, a closed form solution of a shielding position is deduced, and an iterative optimization process of a shielding state point position is eliminated, so that the calculation complexity is reduced, and efficient real-time tracking is realized; and meanwhile, the inherent space-time geometric evolution law of the flexible linear object is utilized to construct a position estimation model which can be directly analyzed and solved, so that the stability risk caused by iterative optimization is avoided, and the reliability and tracking stability of shielding position estimation are remarkably improved.

Disclosed is an apparatus for detecting a linear object, the apparatus including: an ultrasonic sensor including first, second, third and fourth ultrasonic sensor units, each of which is configured to transmit a signal toward an object and receive signals of direct waves and indirect waves reflected from the object; an ultrasonic signal preprocessor configured to preprocess the direct waves and indirect waves received in each of the first, second, third and fourth ultrasonic sensor units and detect a direct wave time-of-flight (TOF) and an indirect wave TOF; a linear object identifier configured to identify a linear object based on at least one of the direct wave TOF and the indirect wave TOF; and a controller configured to control braking of a vehicle based on the linear object.

To provide a linear object holding block which enables a robot hand to hold one linear object easily.SOLUTION: A linear object holding block 72 is capable of holding linear objects and includes: holding parts 74, 77, 79, 81, 84 which hold the linear objects in a manner that the linear objects can be taken out by a robot hand; and restriction parts 75, 82, 85 which restrict movement of the linear objects held by the holding parts in a width direction.SELECTED DRAWING: Figure 8

The invention discloses a ConvNeXt-based deformable linear object semantic segmentation network structure and a ConvNeXt-based deformable linear object semantic segmentation method. The ConvNeXt-based deformable linear object semantic segmentation network structure comprises an input layer; the backbone network comprises five layers of convolutional neural networks of ConvNeXt architecture, each layer comprises a wavelet down-sampling module and a wavelet convolution module, and the backbone network is used for outputting a multi-scale feature map; the input end of the pixel decoder is connected with the output end of the backbone network, and the pixel decoder is used for carrying out fusion and up-sampling on the multi-scale feature map; the input end of the Transform decoder is connected with the output end of the pixel decoder, and the Transform decoder is used for generating an initial binary mask image; and the post-processing module comprises a mask optimization unit and an independent mask generation unit and is used for performing morphological skeletonization processing and intersection processing on the initial binary mask image and finally outputting an independent binary mask of each deformable linear object.

The invention discloses a dynamic etching compensation optimization method for a signal layer circuit. The method comprises the steps of obtaining final compensation parameters, determining compensation intervals, smoothing adjacent compensation intervals and the like. According to the dynamic etching compensation optimization method of the signal layer circuit, the compensation interval of the linear object is obtained in a projection mode, so that the speed and the accuracy of the compensation partition of the linear object are effectively improved, the compensation optimization efficiency and the compensation effect are greatly improved, the production cost is reduced, and meanwhile, the dynamic etching compensation optimization method of the signal layer circuit is suitable for large-scale popularization and application. By adding the transverse stepped lines in the adjacent intervals for transition, the smoothness of compensation is dynamically controlled, and the yield of products is ensured.

To provide a linear object gripping device which is easily reduced in a size thereof and is suitable for picking up one flexible linear object.SOLUTION: A linear object gripping device 10 has: a pair of fingers 20, 31 which can grip a linear object S; a groove 24 which is formed between the pair of fingers, and in which one linear object is fitted; and suction means which sucks the linear object toward the groove. A width of an open part of the groove is larger than a diameter of the linear object in the state that the pair of fingers opened and is smaller than the diameter of the linear object in the state that the pair of fingers closed.SELECTED DRAWING: Figure 2

1. The name of this design product: Linear object retracting and releasing device (I-shaped). 2. Purpose of this design product: used for reeling in and unreeling in wires. 3. The key point of the design of this product lies in its shape. 4. The picture or photo that best illustrates the design points: three-dimensional picture.

A system attached to an aerial vehicle, the system comprising a LiDAR scanner, a camera, and a computing unit, wherein the computing unit is adapted to: identify, from scan data generated by the LiDAR scanner, a non-zero reflection corresponding to a location of a linear object within a free space; determine, based on the non-zero reflection, a distance and angle between the aerial vehicle and the linear object; and control the camera to track and image the linear object using the distance and angle.

We provide an inventory management system that allows users to easily and visually understand the location and quantity of products in stock. [Solution] The inventory management system 10 includes a generation unit 52 for generating a three-dimensional model of a space in which multiple products are arranged, a storage unit 53 for storing product information including the inventory quantity of each product, and a display control unit 54 that displays the three-dimensional model on a display unit 23, indicates the location of a selected product among the multiple products in the three-dimensional model, displays the product information of the selected product together with the three-dimensional model on the display unit 23, displays a linear object on the display unit 23 connecting the location in the three-dimensional model and the product information of the selected product displayed on the display unit 23, and moves the three-dimensional model on the display unit 23 while the linear object connects the location in the three-dimensional model and the product information of the selected product displayed on the display unit 23 when the display unit 23 moves the three-dimensional model.

To provide a method for gripping a thin, flexible and easily deformable linear object by a robot hand even in a situation where it cannot be recognized by a stereo camera.SOLUTION: A laser is emitted from a laser displacement sensor 40 mounted on an arm 31 of a robot 30 along a measurement line crossing a linear object W, and reflected light is received by the laser displacement sensor to measure a distance to a reflection point. Acquiring a reflection-point profile including information on a distance from the laser displacement sensor to the reflection point and a position and a direction of the laser displacement sensor, calculating a corrected reflection-point profile by converting the reflection-point profile into a coordinate system that does not change depending on a posture of the arm, determining a holding position of the line member based on the corrected reflection-point profile, and holding the holding position with the hand 32 attached to the tip of the arm or another hand.SELECTED DRAWING: Figure 1

Embodiments of the present disclosure provide a processing method and device for an augmented reality picture, an electronic device and a storage medium. The method comprises: in response to a rendering trigger request for an augmented reality picture, displaying a three-dimensional model of a linear object corresponding to the augmented reality picture in the augmented reality picture; and in response to a display adjustment operation for the linear object, displaying an adjusted three-dimensional model of the linear object in the augmented reality picture, wherein the display adjustment operation comprises a display position adjustment operation, a display size adjustment operation and / or a display angle adjustment operation. The technical solution of the embodiments of the present disclosure displays a three-dimensional model of a linear object in an augmented reality picture, not only more realistically integrates the linear object into the augmented reality picture, but also improves the stereoscopic and realistic sense of the linear object in the augmented reality picture, and effectively interacts with the linear object according to the individual needs of the user, thereby improving the user experience.

To provide a more inexpensive device for automatically inserting or connecting the tip of a linear object to a target.SOLUTION: A linear object moving device 10 is a device for inserting or connecting a distal end 52 of a linear object 50 into or to a target 60, and includes a triaxial Cartesian coordinate robot 20, a rotational triaxial wrist unit 30 attached to the Cartesian coordinate robot and including a gripping part 34 capable of gripping the linear object, and a three dimensional sensor 40 disposed so as to be capable of measuring a position and a posture of the distal end of the linear object gripped by the wrist unit.SELECTED DRAWING: Figure 1

A method (160) for determining a position of a linear object (170) based on echo information from a vehicle ultrasonic transducer arrangement (140) containing at least one ultrasonic transducer (130) is proposed, which method includes: providing (S3) the echo information describing at least three echo signals, wherein transmission and reception positions of the echo signals are distributed two- or three-dimensionally; and determining (S5, S6, S7) a position of a linear object (170) based on the echo information and an echo-based position determination model of a linear object (170).

The application provides a training method and device of an image segmentation model, a storage medium and an electronic device. The method comprises: labeling each pixel in a sample image according to a pixel type to obtain a sample labeled image; performing background segmentation on the sample image through a first branch network in a neural network to obtain a first segmentation result, wherein the neural network comprises an image segmentation model, and the first branch network is an intermediate model obtained by training the image segmentation model using a first dilated image; and training the image segmentation model using the first segmentation result and a second dilated image to obtain a target image segmentation model, wherein the target image segmentation model has a segmentation accuracy rate for a target object class and a background class in the sample image that is greater than a preset threshold. The above technical solution solves the problem of slow convergence speed of an image segmentation model for linear objects in the training process in related technologies.

The application relates to a linear target recognition method, device, vehicle and equipment. The method comprises the following steps: acquiring a perception image, wherein the perception image comprises a linear target to be recognized; performing discrete sampling on the perception image to obtain a sampling image comprising discrete sampling results; identifying a feature region of the sampling image through a convolutional neural network to obtain position information of a sampling point corresponding to the linear target; and determining a recognition result of the linear target according to the position information of the sampling point corresponding to the linear target. The application can solve the problem that the number of linear target outputs is limited by performing discrete sampling on the perception image. The linear target recognition method has small calculation amount and wide applicability, and can be conveniently extended to detection of other linear objects.