8 results about "Global local" patented technology

The invention discloses an improved agricultural pest larva detection method based on YOLOv8, and relates to the technical field of target detection. Comprising the following steps of: constructing an APL8 (Activated Protein Label 8) data set of eight types of common agricultural pest larvae, and constructing an APL8 (Activated Protein Label 8) data set of the eight types of common agricultural pest larvae; the method comprises the following steps of: constructing a YOLOv8-LQG detection model on the basis of a YOLOv8 model, in combination with a designed LHF (Learnable Hue Filter) module, adding a QDDL (Quadruple Downsampling Description Layer) and a GLCM (Global Local Context Module) which fuses global semantics and local textural features, and constructing the YOLOv8-LQG detection model on the basis of the YOLOv8 model; a YOLOv8-LQG model is trained; and the YOLOv8-LQG model is tested.

The application relates to the technical field of calibration, and discloses a single-station-based rotating shaft calibration method, which comprises the following steps: obtaining a local initial point cloud by rotating scanning a house to be measured by a three-dimensional scanner, and calculating the transformation relationship between the local initial point cloud and a rotating shaft coordinate system to obtain the relative initial external parameter of the three-dimensional scanner and the rotating shaft; extracting a plane feature of the local initial point cloud in the rotating shaft coordinate system to obtain a plane point cloud and an index associated with the plane point cloud, and processing the plane point cloud based on the relative initial external parameter and the index associated with the plane point cloud to obtain a global local point cloud in a global coordinate system; constructing a target optimization function by using the index associated with the plane point cloud and the global local point cloud in the global coordinate system, and iteratively optimizing the relative initial external parameter by using the target optimization function to obtain a relative external parameter. The method provided by the application not only ensures the accuracy of the relative external parameter, but also reduces the time cost and improves the house measurement efficiency.

The invention relates to the technical field of remote sensing images, in particular to a remote sensing image target detection method and system based on a global and local context sensing network, and the method comprises the steps: firstly carrying out the cross fusion of global and local context information of an input remote sensing image through employing a space-channel cross fusion strategy through employing a GLCCFNet, and carrying out the cross fusion of the global and local context information in space and channel dimensions; the GLCCFNet fully considers the complementary characteristics of global and local context information in a channel and a space, so that efficient fusion of the global and local context information is realized; then, a TSFFPN is proposed to fuse the multi-scale feature maps and filter out complex background information, and the TSFFPN not only can reduce information loss of non-adjacent feature layers, but also can automatically select space attention weights required by feature layers with different scales, so that the complex background information is filtered out; and finally, the MSSHead is used for obtaining the features of the extreme length-width ratio target by increasing the receptive field of the fused feature map, and the features of the extreme length-width ratio target are concerned.

The invention discloses a video anomaly detection method and system based on a global local feature double-flow network. The method comprises the following steps: 1.1, loading an original video training and test data set; 1.2, dividing a video frame sequence and constructing a corresponding optical flow frame sequence; 2.1, constructing a video anomaly detection model based on global and local feature fusion; 2.2, the model takes a video frame sequence and an optical flow frame sequence as input of an appearance branch encoder and an input of a motion branch encoder respectively, and appearance feature codes and motion feature codes are extracted respectively; 2.3, restoring the image by the appearance and motion decoder, and generating a prediction frame image matched with the corresponding mode; 2.4, calculating an appearance and motion joint loss function, performing back propagation, and optimizing a codec of a branch; 3.1, dividing the test video into a video frame sequence and an optical flow frame sequence; 3.2, inputting into a trained model to generate a prediction frame image at the next moment; 3.3, calculating abnormal scores of the appearance stream and the motion stream of each test video frame; and 3.4, when the abnormal score exceeds a threshold value, judging the current frame as an abnormal frame.

Systems and methods described herein relate to implementing battery life prediction strategies. In one embodiment, a method includes receiving a first battery dataset containing a first set of battery entries, a first set of historical usage entries, a first set of local parts entries, and a first global parts entry; generating a second battery dataset containing a second set of battery entries and a second set of historical usage entries; generating a second set of local parts entries for the second battery dataset based on comparing the first set of historical usage entries with the second set of historical usage entries; copying the first global parts entry to a second global parts entry of the second battery dataset; and optimizing via one or more global local decomposition transformers the second set of local parts entries and the second global parts entry based on the second set of historical usage entries.

Systems and methods described herein relate to implementing battery life prediction strategies. In one embodiment, a method includes receiving a first battery dataset containing a first set of battery entries, a first set of historical usage entries, a first set of local parts entries, and a first global parts entry; generating a second battery dataset containing a second set of battery entries and a second set of historical usage entries; generating a second set of local parts entries for the second battery dataset based on comparing the first set of historical usage entries with the second set of historical usage entries; copying the first global parts entry to a second global parts entry of the second battery dataset; and optimizing via one or more global local decomposition transformers the second set of local parts entries and the second global parts entry based on the second set of historical usage entries.