57 results about "Scale estimation" patented technology

The invention provides a category-level 6D attitude estimation method based on geometric perception key points, and relates to the technical field of attitude estimation.The method comprises the steps that on the basis that RGB images and point cloud features are fused, a dynamic key point proposing module is designed to generate key points which are consistent in category and geometrically adaptive, so that the adaptability to intra-class deformation and weak texture targets is enhanced; secondly, introducing a spatial geometric attention module to model a spatial structure relationship between key points so as to optimize key point distribution; and finally, point cloud reconstruction and scale regression are simultaneously realized through a geometric perception reconstruction and scale estimation module under the prior condition of no CAD model, and end-to-end optimization is carried out by utilizing multi-task loss. Experimental results on REAL275, CAMERA25 and House Cat6D data sets show that the method provided by the invention is superior to the existing method in multiple indexes such as IoU and attitude precision, and shows stronger generalization and robustness in a real complex scene.

The invention relates to the technical field of engineering surveying and mapping, in particular to a structural point cloud principal axis identification method based on adaptive weighted principal component analysis, which comprises the following steps: acquiring a point cloud data set, initializing a weight set, and performing weighted centralization on the point cloud data set after point cloud downsampling; calculating a weighted covariance matrix according to the point cloud data set after weighted centralization, and obtaining a feature value and a normalized feature vector through the weighted covariance matrix; obtaining a distance measurement set of all point clouds; calculating an iteration weight set; and calculating the eigenvalue of the weighted covariance matrix and outputting an iterative spindle rotation angle. In the application of the method, along with the increase of the complexity of the point cloud, the advantage of adaptive weight estimation is more remarkable than that of common scale estimation, so that the high-precision rapid convergence performance when the common scale estimation is used for processing the uniform and symmetrical point cloud is reserved; and interference of outliers and environmental noise points on main shaft identification can be effectively prevented when highly non-uniform distribution point clouds are processed.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify two-dimensional images via scene-based editing using three-dimensional representations of the two-dimensional images. For instance, in one or more embodiments, the disclosed systems utilize three-dimensional representations of two-dimensional images to generate and modify shadows in the two-dimensional images according to various shadow maps. Additionally, the disclosed systems utilize three-dimensional representations of two-dimensional images to modify humans in the two-dimensional images. The disclosed systems also utilize three-dimensional representations of two-dimensional images to provide scene scale estimation via scale fields of the two-dimensional images. In some embodiments, the disclosed systems utilizes three-dimensional representations of two-dimensional images to generate and visualize 3D planar surfaces for modifying objects in two-dimensional images. The disclosed systems further use three-dimensional representations of two-dimensional images to customize focal points for the two-dimensional images.

An XR system is provided that enhances user interaction within extended reality environments through precise hand scale estimation. The XR system is configured to capture tracking data of a user's hand as the user interacts with a mobile device. Concurrently, the XR system captures pose data of itself and uses the tracking data and the pose data to determine a reference line segment. This segment aids in calculating three-dimensional distances between node pairs of the user's hand. By employing these measurements, the XR system effectively calculates a hand scale factor that is used for accurately integrating the user's hands into an XR user interface.

Examples in the present disclosure relate to temporally sparse scale estimation for object tracking. A computing device detects a current orientation of an object. The computing device determines that a difference between the current orientation and at least one of a plurality of previously detected orientations of the object is less than a threshold value. Each previously detected orientation has a respective scale estimate. In response to determining that the difference is less than the threshold value, the computing device generates an effective scale estimate for the object based on a combination of the respective scale estimates for the plurality of previously detected orientations. Each respective scale estimate contributes to the effective scale estimate according to a respective difference between the current orientation and the previously detected orientation for the respective scale estimate. The computing device tracks a pose of the object based on the effective scale estimate.

Examples in the present disclosure relate to scale estimation for facilitating extended reality (XR) experiences. An image of a hand of a user is captured via one or more optical sensors of an XR device. The image is processed to detect a hand pose relative to the XR device. A hand scale estimate corresponding to the detected hand pose is accessed. The hand scale estimate is one of a plurality of hand scale estimates each uniquely associated with a respective hand pose. The hand scale estimate is applied to generate positional data for one or more features of the hand of the user. The XR device tracks the hand of the user based on the positional data while the user uses the XR device.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify two-dimensional images via scene-based editing using three-dimensional representations of the two-dimensional images. For instance, in one or more embodiments, the disclosed systems utilize three-dimensional representations of two-dimensional images to generate and modify shadows in the two-dimensional images according to various shadow maps. Additionally, the disclosed systems utilize three-dimensional representations of two-dimensional images to modify humans in the two-dimensional images. The disclosed systems also utilize three-dimensional representations of two-dimensional images to provide scene scale estimation via scale fields of the two-dimensional images. In some embodiments, the disclosed systems utilizes three-dimensional representations of two-dimensional images to generate and visualize 3D planar surfaces for modifying objects in two-dimensional images. The disclosed systems further use three-dimensional representations of two-dimensional images to customize focal points for the two-dimensional images.

The invention relates to the technical field of ore body three-dimensional positioning prediction, and discloses a hidden ore body three-dimensional positioning prediction method based on artificial intelligence, and the method comprises the steps: firstly processing geological data through data preprocessing and feature extraction, and then carrying out the training and prediction through a constructed neural network model, and three-dimensional positioning of the concealed ore body is realized. The method comprises the steps of data normalization, missing value filling, feature extraction, neural network model construction, gradient descent updating, three-dimensional coordinate conversion, ore body scale estimation and the like. Through prediction precision evaluation and model optimization, the accuracy and efficiency of concealed ore body positioning are improved. The invention provides an efficient and accurate three-dimensional positioning prediction method for mineral resource exploration, and the method has a wide application prospect.

The invention provides a method and a device for constructing a three-dimensional simulation scene driven by multi-modal data, and relates to the technical field of computer vision. Comprising the following steps: acquiring and preprocessing multi-modal data so as to encode the multi-modal data into a joint embedded vector; generating a scene layout image according to the joint embedded vector and a condition image generation model, and segmenting the scene layout image according to a semantic segmentation expert model and a soft voting mechanism to determine a semantic segmentation result; performing boundary extraction on the semantic segmentation result to determine a standardized scene semantic graph; selecting an alternative digital asset set based on a category label corresponding to the semantic segmentation result and a simulation engine digital asset library; inputting the scene layout image into a scale estimation model to output a scene map scale; and generating a three-dimensional simulation scene according to the standardized scene semantic graph, the alternative digital asset set, the scene map scale and a two-dimensional layout algorithm. According to the invention, the scene synthesis effect from the multi-modal intention to the landing simulation scene is improved.

Machine Learning based scaling estimation in the art mostly used a complex guided wave setup. Further, selection of features for ML is critical to the predict significant points on the US signal response providing appropriate but minimum features that capture maximum scaling characteristics, thus keeping the ML features minimal to provide time and resource efficient computation. A method and system for estimation of scaling in objects by processing ultrasound responses using ML is disclosed. The system uses low-voltage pulse packets to generate ultrasonic (US) waves, capable of penetrating metal structures, using an economically repurposed piezoelectric transducer. These US signal response from the object-scaling interface and scaling-fluid interface, is captured using the repurposed piezoelectric transducer and processed to generate the envelope of the US signal response, and unique 42 features are extracted. A pretrained ML model processes the features to estimate various levels of scaling present in the object.

The application provides a class-level 6D pose estimation method based on geometric perception key points, and relates to the technical field of pose estimation. On the basis of fusing RGB image and point cloud features, a dynamic key point proposal module is designed to generate class-consistent and geometric-adaptive key points to enhance the adaptability to intra-class deformation and weak texture targets. Secondly, a spatial geometry attention module is introduced to model the spatial structure relationship between key points to optimize the key point distribution. Finally, through the geometric perception reconstruction and scale estimation module, point cloud reconstruction and scale regression are simultaneously realized under the condition of no CAD model prior, and end-to-end optimization is carried out by using multi-task loss. The experimental results on REAL275, CAMERA25 and HouseCat6D datasets show that the method of the application is superior to existing methods in IoU and pose accuracy indexes, and has stronger generalization and robustness in real complex scenes.

The present application relates to the field of engineering surveying and mapping technology, and particularly relates to a structure point cloud principal axis identification method based on adaptive weighted principal component analysis, which comprises the following steps: obtaining a point cloud data set, initializing a weight set, and weighting and centralizing the point cloud data set after point cloud downsampling; calculating a weighted covariance matrix according to the weighted and centralized point cloud data set, and obtaining eigenvalues and normalized eigenvectors through the weighted covariance matrix; obtaining a distance measurement set of all point clouds; calculating an iterative weight set; and calculating the eigenvalues of the weighted covariance matrix to output an iterative principal axis rotation angle. In application, with the increase of point cloud complexity, the advantage of adaptive weight estimation is more and more significant compared with the common scale estimation, which not only retains the high-precision fast convergence ability of the common scale estimation when processing uniform and symmetric point clouds, but also effectively prevents the interference of outlier points and environmental noise points on the principal axis identification when processing highly non-uniformly distributed point clouds.

The present invention discloses a remote sensing image segmentation method that combines a quadtree structure with scale estimation. The method specifically includes remote sensing image preprocessing, quadtree partitioning, spatial segmentation scale estimation, attribute segmentation scale estimation, fractal network evolution algorithm segmentation, merging segmentation results, and accuracy verification. To address the image over-segmentation, under-segmentation, and mis-segmentation phenomena caused by the scale characteristics of different land types, the present invention uses the FNEA algorithm as a basis, a quadtree structure, and a scale estimation method based on spectral spatial statistics to achieve remote sensing image segmentation that includes multiple land feature features. This effectively improves segmentation accuracy and avoids the problems of linear land feature extraction interruption and classification errors caused by mis-segmentation and under-segmentation. This method is of great significance for remote sensing image classification and information extraction.

The invention provides an optimization and acceleration method and system for a visual inertial navigation system of an unmanned aerial vehicle based on heterogeneity. The method comprises the following steps: S1, outputting a reliable feature point set which is a direct input object for performing initial scale estimation in S2; and the feature points established in the step S1 become seed features tracked in the step S3. And S3, based on the position of the feature point determined in the S1, performing tracking by using a hardware acceleration optical flow in a subsequent frame. And the tracking result of the S3 enters the S4 for bidirectional optical flow consistency check, feature points which fail in tracking are eliminated, and a high-reliability feature trajectory is output. S4, taking the verified feature track as input of S5 feature management, and performing dynamic density control and quality evaluation. And S5, taking the managed feature distribution as a candidate set for feature selection in S6, and selecting an optimal subset for back-end optimization based on information gain evaluation in S6. S7, an adaptive mechanism adjusts the selection strategy in the step S6; and S8, overall system parameters are adjusted according to the state of the computing resource.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify two-dimensional images via scene-based editing using three-dimensional representations of the two-dimensional images. For instance, in one or more embodiments, the disclosed systems utilize three-dimensional representations of two-dimensional images to generate and modify shadows in the two-dimensional images according to various shadow maps. Additionally, the disclosed systems utilize three-dimensional representations of two-dimensional images to modify humans in the two-dimensional images. The disclosed systems also utilize three-dimensional representations of two-dimensional images to provide scene scale estimation via scale fields of the two-dimensional images. In some embodiments, the disclosed systems utilizes three-dimensional representations of two-dimensional images to generate and visualize 3D planar surfaces for modifying objects in two-dimensional images. The disclosed systems further use three-dimensional representations of two-dimensional images to customize focal points for the two-dimensional images.

The application belongs to the technical field of target tracking, and discloses a single-target long-time tracking method based on a deep neural network, which comprises the following steps: acquiring a tracking target image and a tracking search area image, building a feature extraction network based on a twin network, and performing feature extraction; in the tracking process, a target image is selected by a PNR score to be added to a template library and is sent to a model learning network to perform model learning and updating, a model target feature map and a search feature map are convolved to obtain a response map, the response map is sent to a scale regression network to obtain a scale position score map, and a target scale is acquired; when the target is occluded or lost, target recapture is performed; in the tracking process, information of the target and a suspected target is recorded, and when an interference object appears, a tracking position that is most matched with a historical track is selected. The application reduces model drift phenomenon, makes scale estimation more stable and accurate, and makes the algorithm suitable for long-time target tracking by using a re-search mechanism and an anti-interference matching technology.

Examples in the present disclosure relate to temporally sparse scale estimation for object tracking. A computing device detects a current orientation of an object. The computing device determines that a difference between the current orientation and at least one of a plurality of previously detected orientations of the object is less than a threshold value. Each previously detected orientation has a respective scale estimate. In response to determining that the difference is less than the threshold value, the computing device generates an effective scale estimate for the object based on a combination of the respective scale estimates for the plurality of previously detected orientations. Each respective scale estimate contributes to the effective scale estimate according to a respective difference between the current orientation and the previously detected orientation for the respective scale estimate. The computing device tracks a pose of the object based on the effective scale estimate.

The application provides a visual positioning scale compensation Kalman filtering method, which realizes online real-time estimation and dynamic compensation of scale error by fusing height information, kinematic constraint and visual observation through explicitly taking scale error factor as a state variable in a state space model of extended Kalman filtering (EKF). The technical scheme of the application is applied to solve the technical problems of unstable scale estimation and large accumulated error of pose in the prior art visual inertial odometer (VIO) system when facing severe motion, visual degradation or low-cost IMU zero bias.

The application discloses a geometric consistency and instrument anchoring three-dimensional reconstruction and polyp measurement method of an endoscope. The method first projects monocular image features to a Riemannian manifold space through a geometric consistency perception module, generates a high-fidelity relative depth map by using a manifold-constrained latent diffusion model, and maintains cross-domain structural consistency. Secondly, by using a scale estimation module based on instrument anchoring, biopsy forceps in a field of view are automatically identified and key points are extracted, a PnP solver and a Bayesian regression are used to solve six degrees of freedom of the instrument, and finally, geometric constraint equations are constructed based on the known physical size of the biopsy forceps, a global absolute scale factor is solved, and the relative depth map is converted into a metric three-dimensional point cloud. A few-sample linear calibration mechanism is introduced, and linear deviation caused by domain offset can be eliminated by using a small amount of clinical frames. According to the application, sub-millimeter polyp size measurement can be realized under monocular endoscopy by using conventional surgical instruments without additional hardware.

Embodiments include methods for synchronizing an augmented reality (AR) object placed in a 3D mesh onto a video feed. A computing device may first receive a video feed including a sequence of frames of a scene and depth or motion data captured by a camera. The computing device may generate a three-dimensional (3D) mesh based on the depth or motion data. The computing device may texture the 3D mesh to create a 3D model. Upon performing object recognition, the computing device may identify anchor points in the 3D model and anchor points in the video feed. The anchor points are used to calculate the location of the AR object.

The application discloses a double-parameter HP filtering denoising algorithm and process for infrared spectrum signal processing, wherein upper and lower limits Lamda_1 and Lamda_2 of parameter scale estimation are given in advance, two filtered signals are obtained by performing traditional HP filtering respectively, the two filtered signals are introduced to comprehensively construct a new HP filtering expression, HP filtering is performed under the new comprehensive reconstructed HP filtering expression, and the parameter value when the expression takes a minimum value is taken as an output result in the range of [Lamda_1, Lamda_2] for optimization searching. The local extreme value is searched in the two parameter estimation ranges, and HP filtering is performed under the new expression. Since the application constructs the HP filtering expression of the comprehensive parameter upper and lower limits and performs the optimization processing in the parameter upper and lower limits, the scale estimation of the signal noise can be better, the problem that the uncertain selection of the parameter in the HP filtering denoising process leads to the difficulty in grasping the scale of the high-frequency noise denoising can be solved, and better denoising effect can be obtained.

The invention relates to the technical field of equipment predictive maintenance, in particular to an industrial equipment residual life prediction method based on an SRenbPI algorithm, which comprises the following steps: preprocessing industrial equipment source domain data, including constructing a heteroscedasticity noise simulation data set and introducing real industrial sensor data, and setting standardized input through normalization, sliding window construction and label; in the training stage, a bootstrap sampling strategy is adopted to construct an integrated regression model, noise influence is quantized through a scaling residual formula, and a residual set is formed; in the prediction stage, noise scale estimation and residual quantile are combined to generate an unequal-width prediction interval, meanwhile, a dynamic updating mechanism is introduced, a residual set is updated in real time to adapt to data distribution changes after a plurality of test samples are processed every time, and the model does not need to be trained again. According to the method, the adaptive defect of a traditional static residual set in industrial equipment full-life-cycle monitoring is effectively overcome, and the prediction interval precision and real-time performance are remarkably improved.

The application discloses a scale adaptive target tracking method in an unmanned aerial vehicle scene, and is used for solving the problems of tracking frame drift and scale estimation inaccuracy caused by the change of visual angle, target deformation and occlusion in the aerial photography of the unmanned aerial vehicle. The core steps include: obtaining a first frame of a video and establishing an initial template of a target; determining a search area of a current frame based on the position of the target in a previous frame, extracting features and calculating a response map to obtain a preliminary center position of the target; innovatively taking the preliminary position as the center to construct a two-dimensional candidate frame set containing different scale factors and different length-width ratio factor combinations; performing feature matching and response value calculation on each candidate frame one by one, and selecting the optimal one as the final position and boundary frame of the target in the current frame to realize the adaptive joint estimation of the scale and the ratio; and the application significantly improves the precision and robustness of target tracking in the unmanned aerial vehicle scene through the scale-ratio joint adaptive estimation and four-dimensional confidence double-branch processing mechanism.

The application relates to the technical field of policy optimization, in particular to an advertisement launching policy optimization method based on AI Agent label understanding, which comprises the following steps: receiving advertisement product information, utilizing a large language model to perform semantic analysis on the information to generate a structured product semantic portrait, and generating an advertisement policy theme according to the structured product semantic portrait; mapping the policy theme and user labels to a high-dimensional semantic vector space to perform K-NN retrieval, utilizing an AI Agent to deduce a Boolean logic connection relationship among label nodes, and constructing an initial virtual policy topology; establishing a policy evolution closed loop based on launching constraints, calling a real-time scale calculation engine to analyze the policy and calculate audience scale, performing topology variation and redundancy pruning on the virtual policy topology according to the deviation direction of scale estimation data and a preset interval, and dynamically adjusting the logic connection to output a final policy. The application realizes optimization of an advertisement policy through AI Agent label understanding and real-time crowd estimation.

An XR system is provided that enhances user interaction within extended reality environments through precise hand scale estimation. The XR system is configured to capture tracking data of a user's hand as the user interacts with a mobile device. Concurrently, the XR system captures pose data of itself and uses the tracking data and the pose data to determine a reference line segment. This segment aids in calculating three-dimensional distances between node pairs of the user's hand. By employing these measurements, the XR system effectively calculates a hand scale factor that is used for accurately integrating the user's hands into an XR user interface.

Examples in the present disclosure relate to scale estimation for facilitating extended reality (XR) experiences. An image of a hand of a user is captured via one or more optical sensors of an XR device. The image is processed to detect a hand pose relative to the XR device. A hand scale estimate corresponding to the detected hand pose is accessed. The hand scale estimate is one of a plurality of hand scale estimates each uniquely associated with a respective hand pose. The hand scale estimate is applied to generate positional data for one or more features of the hand of the user. The XR device tracks the hand of the user based on the positional data while the user uses the XR device.

The invention discloses an engineering field three-dimensional reconstruction physical scale estimation method with reference to a standard component, and belongs to the technical field of building engineering digitization, three-dimensional reconstruction and intelligent construction, and the method comprises the steps: constructing a three-dimensional reconstruction physical scale estimation platform; a multi-view image sequence containing a standard component is collected, a scale-free three-dimensional point cloud and a camera pose are obtained through structure self-motion reconstruction SfM and multi-view stereo matching MVS reconstruction, a point cloud subset of the standard component is detected and positioned, a reconstruction scale size is obtained through geometric fitting, a scale factor is calculated in combination with a real physical size, and a three-dimensional point cloud is obtained. And carrying out global scaling on the scale-free model to finally obtain a real physical scale three-dimensional model. The method is suitable for various image acquisition devices and construction scenes, the scale recovery precision reaches the millimeter-to-centimeter level, the robustness is high, and the method can be widely applied to engineering measurement, BIM model alignment, digital twinning construction and other scenes and has remarkable engineering application value.

The invention discloses a gray measurement expert knowledge generalization model construction method, which comprises the following steps of: constructing a dual-task deep learning model sharing an encoder by synchronously acquiring quantitative scores and visual focus data of experts, so that the dual-task deep learning model learns and predicts expert scores and regions of interest at the same time; the method comprises the following steps of: establishing a deep model, migrating implicit knowledge of the deep model to an interpretable shallow model by utilizing knowledge distillation, extracting dominant and readable discrimination rules, and finally establishing a man-machine collaborative closed loop: submitting highly uncertain samples of the rules and the model to an expert for auditing and correcting, and utilizing feedback iteration to finely adjust the model to realize continuous optimization of performance and interpretability. According to the method, recessive and intuitive knowledge and dominant knowledge of experts are effectively fused, the generalization performance, robustness and interpretability of the model are remarkably improved, and a solid technical support is provided for high-precision and high-reliability automatic gray scale measurement and calculation.

The application discloses an infrared small target detection method based on space-time context perception and compact geometric representation, comprising the following steps: acquiring three adjacent images in a sequence of infrared images to be detected; inputting a pre-trained small target detection model to output a center heat map, a center offset and an effective radius prediction result, and completing positioning and scale estimation of the infrared small target; the small target detection model is used to extract multi-time space features through a backbone network and a feature pyramid sharing weights, obtain space-time representation features suitable for infrared small target detection by introducing space-time context perception information and constructing a time domain difference enhancement and global gating adjustment mechanism, and realize direct prediction of the center position and the effective radius of the small target in combination with a decoupled geometric parameter prediction network. The infrared small target detection task is modeled as target center position and effective radius prediction, so that effective detection is realized while reducing model calculation complexity and labeling cost.

The application provides a scale-invariant image zoom scale mode perception method, comprising the following steps: training a data set of standard scale samples to obtain a trained convolutional neural network; zooming in and out each random scale sample in a test set to form a multi-layer image pyramid composed of multiple scale samples; using the trained convolutional neural network to construct a set of twin convolutional neural networks matched with the number of layers of the multi-layer image pyramid, inputting the multiple scale samples into the set of twin convolutional neural networks for parallel reasoning to obtain a multi-channel classification score matrix; and classifying and estimating the scale according to the classification score matrix. The method has good scale invariance, can classify and estimate the unknown scale at the same time, can be applied to scale-invariant mode recognition tasks of medical images, remote sensing images and the like, and has wider applicability.