Abstract: A method for measuring cargo, performed by a computing device including a processor and a memory, may comprise connecting to a multi-sensor-based cargo measuring device comprising a plurality of 3D depth sensors, obtaining a plurality of raw point clouds corresponding to the cargo measured by the multi-sensor-based cargo measuring device, generating a single aligned point cloud by matching the plurality of raw point clouds to a single space, generating a refined point cloud by performing preprocessing on the aligned point cloud, constructing a K-dimensional tree (KD-Tree) space based on a mesh vertex, placing a ground-truth point cloud in a KD-Tree space, calculating of adjacent points and distances between the adjacent points and determining an uncertainty value, generating a heatmap mesh as a final model based on the uncertainty value, and performing a measurement with respect to the cargo by using the final model.

Abstract: A method for measuring cargo may comprise connecting to a multi-sensor-based cargo measuring device provided with a plurality of 3D depth sensors, obtaining a plurality of raw point clouds corresponding to the cargo measured by the multi-sensor-based cargo measuring device, generating a single aligned point cloud by matching the plurality of raw point clouds to a single space, generating a refined point cloud by performing preprocessing on the aligned point cloud, constructing a K-dimensional tree (KD-Tree) space based on a mesh vertex, with respect to the mesh reconstructed from refined point cloud (a 3D mesh), placing a colored point cloud in the KD-Tree space, calculating adjacent points and distances between the adjacent points, generating a final model by performing color matching by collecting colors of points adjacent to the mesh vertex, and performing a measurement with respect to the cargo by using the final model.

Abstract: A method for measuring cargo, comprising connecting to a multi-sensor-based cargo measuring device provided with 3D depth sensors, obtaining a plurality of raw point clouds corresponding to the cargo measured by the multi-sensor-based cargo measuring device, generating a single aligned point cloud by matching the plurality of raw point clouds to a single space, determining a corner point among points forming the aligned point cloud, determining a point other than the corner point among the points forming the aligned point cloud as a sampling-target point, performing down-sampling on the sampling-target point, generating a refined point cloud, by combining the point for which the down-sampling was performed and corner point, reconstructing a final model by inputting the refined point cloud into a deep learning model, and causing the deep learning model to predict a 3D mesh, and performing a measurement with respect to the cargo by using the final model.

Abstract: A method for measuring cargo, comprising obtaining a plurality of raw point clouds corresponding to the cargo measured by the multi-sensor-based cargo measuring device, generating a single aligned point cloud by matching the plurality of raw point clouds to a single space, determining an operation mode for distinguishing the cargo and a bottom surface supporting the cargo in the aligned point cloud, performing a primary filtering for distinguishing the cargo and the bottom surface from the aligned point cloud, performing a secondary filtering for distinguishing the cargo and the bottom surface in a region where the cargo and the bottom surface are in contact, generating a refined point cloud, reconstructing a final model by inputting the refined point cloud into a deep learning model, and causing the deep learning model to predict a 3D mesh, and performing a measurement with respect to the cargo by using the final model.

Abstract: A method for measuring cargo may include obtaining a plurality of raw point clouds corresponding to the cargo measured by the multi-sensor-based cargo measuring device, generating a single aligned point cloud by matching the plurality of raw point clouds to a single space, determining an operation mode for removing an outlier from the aligned point cloud, setting a plurality of sections in the aligned point cloud, depending on the determined operation mode, determining of a first section including the largest number of points, from among the plurality of sections, and determining a second section adjacent to the first section and having no point, generating a refined point cloud from which the outlier is removed, reconstructing a final model by inputting the refined point cloud into a deep learning model, causing the deep learning model to predict a 3D mesh, and performing a measurement with respect to the cargo.

Abstract: A cargo measuring system includes a multi-sensor-based cargo measurer comprising a plurality of heterogeneous sensors, and a cargo measuring device. The cargo measuring device is configured to receive, from the multi-sensor-based cargo measurer, 3D information of a cargo measured by the multi-sensor-based cargo measurer, and perform measurement of the cargo based on the 3D information, where the 3D information includes a first raw point cloud, a second raw point cloud, a third raw point cloud, and a fourth raw point cloud, each corresponding to a respective side surface of the cargo, and obtained by four 3D depth sensors installed on the multi-sensor-based cargo measurer, the sensors being positioned to capture the side surfaces of the cargo at 90-degree intervals.

Abstract: A cargo measuring method, performed by a computing device including one or more processors and one or more memories, can include: connecting to a multi-sensor-based cargo measurer comprising a plurality of 3D depth sensors, obtaining a plurality of raw point clouds associated with a cargo measured by the multi-sensor-based cargo measurer, matching the plurality of raw point clouds into a single space to thereby create a single aligned point cloud, performing preprocessing on the aligned point cloud to thereby create a refined point cloud, inputting the refined point cloud into a deep learning model and reconstructing a final model by causing the deep learning model to predict a 3D mesh, and performing measurement of the cargo based on the final model.

Abstract: An object extraction method using an object extraction system that includes a pre-processing module, an edge module, and an optimization module, and the object extraction method includes: generating an edge depth map with respect to a 2D image by pre-processing a 2D image where a target object to be data-labeled may be included, by the pre-processing module; extracting a target edge with respect to the target object from 3D data where the target object may be included, by the edge module; and optimizing the generated target edge to a final edge to match the generated edge depth map, by the optimization module.

Abstract: A logistics loading work control system includes a server configured to mirror facility data of an actual logistics facility and a virtual environment, for example, based on a digital twin. The system also includes a packaging simulator configured to calculate cargo information corresponding to a loading area received from the server through a loading algorithm and predict priority cargo information in a sequence of highest loading efficiency with respect to a current work-target cargo container, and a worker terminal configured to assist a logistics loading work of a worker by deriving a scenario branch-type loading sequence and an expected loading rate based on a logistics loading situation inside the loading area through the packaging simulator and providing a recommended cargo logic according to the derived result.

Abstract: A durability evaluation system includes a collector collecting first data for each of meshes forming a vehicle body of a target vehicle, and second data for each mesh forming a vehicle body of a virtual vehicle in which a material or a thickness of the target vehicle is adjusted and a controller determining input data for first meshes based on the lifetime data matched to each row for the meshes of data obtained by preprocessing the first data and the second data, extracting feature data including features representing the stress analysis data for each of the first meshes and each of second meshes adjacent to the first meshes and features representing a shape of each of the first meshes, generating first synthesized data based on the feature data, and learning a method of predicting the lifetime data representing durability performance based on the feature data and the first synthesized data.

Abstract: A durability evaluation method performed by a processor including filtering, by a data filter, a first data set for each of a plurality of meshes of a target vehicle and a second data set with an adjusted material or thickness of the target vehicle to output data, determining, by a binary classifier, the first synthetic data for indicating durability performance less than a predetermined threshold value from feature data from the filtered data and the first synthetic data, generating, by a probability inferring unit, second output data for representing the probability belonging to each of a predetermined number of durability probability sections, from the durability performance of the first output data and the second synthetic data, and determining a plurality of hyperparameter values constituting each of the data filter, the binary classifier, and the probability inferring unit.

Abstract: A durability evaluation system includes a collector collecting first data for each of meshes forming a vehicle body of a target vehicle, and second data for each mesh forming a vehicle body of a virtual vehicle in which a material or a thickness of the target vehicle is adjusted and a controller determining input data for first meshes based on the lifetime data matched to each row for the meshes of data obtained by preprocessing the first data and the second data, extracting feature data including features representing the stress analysis data for each of the first meshes and each of second meshes adjacent to the first meshes and features representing a shape of each of the first meshes, generating first synthesized data based on the feature data, and learning a method of predicting the lifetime data representing durability performance based on the feature data and the first synthesized data.

Abstract: An object extraction method using an object extraction system that includes a pre-processing module, an edge module, and an optimization module, and the object extraction method includes: generating an edge depth map with respect to a 2D image by pre-processing a 2D image where a target object to be data-labeled may be included, by the pre-processing module; extracting a target edge with respect to the target object from 3D data where the target object may be included, by the edge module; and optimizing the generated target edge to a final edge to match the generated edge depth map, by the optimization module.

Abstract: An adverse situation warning device includes a route data storing portion for storing a first route data of a host vehicle including a first geohash corresponding to a moving route of the host vehicle, a route comparing portion for receiving a second route data of a remote vehicle including a second geohash corresponding to a moving route of the remote vehicle, and for comparing the geohash of the first route data and the geohash of the second route data, and a warning output portion for outputting an adverse situation warning based on the comparison result.

Abstract: An adverse situation warning device includes a route data storing portion for storing a first route data of a host vehicle including a first geohash corresponding to a moving route of the host vehicle, a route comparing portion for receiving a second route data of a remote vehicle including a second geohash corresponding to a moving route of the remote vehicle, and for comparing the geohash of the first route data and the geohash of the second route data, and a warning output portion for outputting an adverse situation warning based on the comparison result.