Abstract: An in-oil autonomous operation detection robot of a storage tank bottom plate and an autonomous operation detection method are provided. The in-oil autonomous operation detection robot of the storage tank bottom plate includes a motion module, a positioning and attitude recognition module, an obstacle avoidance module, a detection module and a control module, wherein the motion module is configured to adjust and control a motion direction, a speed and an attitude of the robot under the control of the control module; the positioning and attitude recognition module is configured to recognize a position and an attitude of the robot under the control of the control module; the obstacle avoidance module is configured to avoid obstacles under the control of the control module; and the detection module is configured to detect corrosions of the bottom plate.

Abstract: Disclosed is interactive labeling of a 4D dynamic object based on time series data, which aims at time series-related point cloud dynamic object data. Multi-frame local point clouds in the same time series are transformed into the same global coordinate system with corresponding poses to obtain global point clouds in the same time series are obtained, which clearly shows the moving trajectory of the dynamic object. Taggers can label key 3D boxes based on the moving trajectory of the dynamic object, and automatically generate 3D prediction boxes of other frames based on these key 3D boxes, which significantly reduces the number of frames that need manual operation and solves the problem that 3D prediction boxes generated based on deep learning model are inaccurate and efficiency can hardly be improved.

Abstract: Disclosed is interactive labeling of a 4D dynamic object based on time series data, which aims at time series-related point cloud dynamic object data. Multi-frame local point clouds in the same time series are transformed into the same global coordinate system with corresponding poses to obtain global point clouds in the same time series are obtained, which clearly shows the moving trajectory of the dynamic object. Taggers can label key 3D boxes based on the moving trajectory of the dynamic object, and automatically generate 3D prediction boxes of other frames based on these key 3D boxes, which significantly reduces the number of frames that need manual operation and solves the problem that 3D prediction boxes generated based on deep learning model are inaccurate and efficiency can hardly be improved.

Abstract: A leakage magnetic field graph testing method for in-service cable damage includes: acquiring multi-channel detection signals corresponding to respective testing units in a testing probe; slicing the multi-channel detection signals acquired at different detection positions to acquire sliced detection data at a plurality of detection positions in an axial direction of a cable under test; mapping the sliced detection data to a pre-constructed graph structure to acquire sliced graphs, wherein the graph structure is established on the basis of spatial distribution of the testing units in the testing probe, the testing units serving as nodes, and adjacency relationships between the testing units in the testing probe serving as edges; and determining a frequency spectrum of a graph signal in each sliced graph by means of a graph Fourier transform.

Abstract: The present invention discloses a pavement element annotation method for point cloud data with fusion of height, which comprises the following steps: constructing all single-frame point clouds into a joint point cloud in a global coordinate system based on the pose of each frame of single-frame point cloud; removing dynamic objects in the joint point cloud to obtain a static joint point cloud; transforming the static joint point cloud into an overhead image; pre-annotating the pavement elements in an overhead image by using the pavement element pre-annotation model; modifying the pre-annotated result to obtain an overhead image annotation; based on a ball query algorithm and a ground point algorithm, establishing a transformational relation between the pixels of the overhead image and the points of the static joint point cloud; transforming the overhead image annotation into a static joint point cloud annotation based on the transformational relation.

Abstract: A method and an apparatus for 4D road scene annotation based on time series data includes: obtaining point cloud data and multi-view 2D image data of a same road scene at a same time series, as well as a rotation matrix R and a translation vector t obtained through sensors; merging all the point cloud data into a combined point cloud using the rotation matrix R and the translation vector t; performing dynamic and static object annotation and lane line annotation on the combined point cloud in the 4D road scene; and mapping annotation information in the 4D road scene to all the 2D image data at the time series based on camera parameter information, to obtain annotation information on 2D images. The method fully utilizes time series information, avoiding the loss of important information.

Abstract: Provided is a test sensor using a magnetostrictively induced guided wave based on an open magnetic circuit, comprising an excitation coil, a receiving coil and a magnetic device, the magnetic device comprises multiple test modules circumferentially and uniformly disposed thereon so as to be absorbed to the outer side of a to-be-tested slender component, each test module comprises a housing, a permanent magnet and a magnetic plate, two adjacent housings are connected to each other via an adjusting device, the excitation coil and the receiving coil are disposed in the vicinity of the test module, and are coaxially fit on the outer side of the to-be-tested slender component, the excitation coil operates to generate induced voltage in the receiving coil after the sinusoidal alternating current is input, and a computer can determine whether a defect occurs in the to-be-tested slender component after receiving the induced voltage.

Abstract: A method for denoising magnetostrictive guided wave detection signals to improve detection accuracy. The method includes forming a matrix A by using the signals; performing a singular value decomposition on the matrix A to obtain a singular matrix B including a plurality of eigenvalues; setting eigenvalues in the singular matrix B that are smaller than the median to zero to obtain a matrix C; performing an inverse transformation of the singular value decomposition on the matrix C to obtain a matrix D; and determining the denoised signals according to the matrix D.

Abstract: Provided is a test sensor using a magnetostrictively induced guided wave based on an open magnetic circuit, comprising an excitation coil, a receiving coil and a magnetic device, the magnetic device comprises multiple test modules circumferentially and uniformly disposed thereon so as to be absorbed to the outer side of a to-be-tested slender component, each test module comprises a housing, a permanent magnet and a magnetic plate, two adjacent housings are connected to each other via an adjusting device, the excitation coil and the receiving coil are disposed in the vicinity of the test module, and are coaxially fit on the outer side of the to-be-tested slender component, the excitation coil operates to generate induced voltage in the receiving coil after the sinusoidal alternating current is input, and a computer can determine whether a defect occurs in the to-be-tested slender component after receiving the induced voltage.

Abstract: A method for processing magnetostrictive guided wave detection signals, including: 1) obtaining an analysis signal by capturing an original magnetostrictive guided wave detection signal; 2) performing band-pass filtering on the analysis signal to obtain a signal, and initializing i to 0; 3) obtaining a group of signals x(i), x(i+1), . . . , x(i+M?1) using a rectangular window with a width of M; 4) forming a matrix A; 5) performing singular value decomposition on the matrix A to obtain a singular matrix B; 6) setting eigenvalues in the matrix B smaller than the median to 0 to obtain a matrix C, and performing inverse singular value transformation on the matrix C to obtain a matrix D; 7) recovering a group of processed signals from the matrix D and calculating energy z of the group of processed signals; and 8) setting i to (i+1) and repeating steps 3)-7) until i=N+1?M.