3results about How to "Achieve completeness" patented technology

The application discloses a large field of view structure image classification method based on continuous deformation modeling, mainly solves the problem that the existing dynamic deformation modeling is insufficient in integrity, resulting in poor classification effect. Including: 1) constructing progressive structure similar samples, realizing self-learning adaptation of non-rigid structure intra-class change; 2) constructing a multi-branch attention representation learning network composed of three parallel branches of double-flow spatial attention, adaptive fusion channel attention and original features; 3) configuring a multi-branch selection fusion module based on the learning network, and constructing a continuous deformation feature extraction network; 4) inputting the continuous deformation fusion features output by the extraction network into a classifier to realize self-learning classification of aurora large field of view structure images. The application can automatically learn the structure change rule under the condition of limited labeling or unlabeled, significantly improve the discrimination ability and generalization performance of complex deformation structure, and is suitable for aurora, cloud cluster and other non-rigid deformation image analysis.

The application discloses a kind of automatic expansion and contraction and divides the method, device, equipment and storage medium of heap, the method is applied to packaging substrate manufacturing process, before mechanical drilling process, the method includes: product is fed and the product information of current product is obtained;Based on the product information, determine whether the current product is first batch product;If it is determined to be, then based on the preset basic pre-expansion value, perform X-ray drilling target operation and obtain actual expansion and contraction value, according to the deviation of actual expansion and contraction value and basic pre-expansion value, determine whether it needs to be divided into piles;If it is determined to be not, then based on historical expansion and contraction data, perform X-ray drilling target operation and determine whether it needs to be divided into piles;According to the product is divided into piles according to the result of heap, for subsequent mechanical drilling process selects corresponding expansion and contraction data for processing.The application can realize the intelligent management of expansion and contraction and divide the heap, improve efficiency and yield.

The invention discloses a pull sling response signal processing method and device based on empirical mode decomposition, a medium and equipment, and belongs to the technical field of bridge operation and maintenance. The method comprises the following steps: collecting real-time data of a target signal according to a preset moving window length to obtain current moving window data; performing signal decomposition on the current moving window data based on an empirical mode decomposition method to obtain a current moving window decomposition signal; and eliminating data breakpoints of overlapped parts by adopting an overlapped region linear weighted fusion method, taking decomposition results obtained in respective moving windows for non-overlapped parts, and combining the decomposition results to obtain smooth and continuous corrected decomposition signals. According to the method, the calculated amount pressure of long-time sequence signal loading and processing can be avoided, boundary data distortion of segmented processing is eliminated, the continuity of each signal component after decomposition is guaranteed, the completeness of signal characteristics is guaranteed, and the reconfigurability of the signals is reserved.