30 results about "Overcast" patented technology

The invention discloses an environment monitoring big data integration system based on computer services, comprising a temperature information collection module, a weather information collection module, an illuminance collection module, a wind information collection module, an air information collection module, a data receiving module, a data processing module, a data integration module, a master control module and an information sending module; the temperature information collection module is used to collect temperature information, the temperature information includes indoor temperature information and outdoor temperature information, and the weather information collection module is connected to the Internet to obtain weather information from the network , the weather information includes: precipitation, sunny days and cloudy days; the illuminance collection module is used to collect outdoor illuminance, and the wind information collection module is used to collect wind information. The beneficial effects of the present invention are: data can be more comprehensive and faster At the same time, prompt information can be integrated to meet the different needs of users.

The invention relates to a household ventilating and rain-shielding device, and belongs to the technical field of intelligent household products. The household ventilating and rain-shielding device comprises a wall fixing bottom plate, a ball screw mechanism, a linear guide rail mechanism, a nut sliding plate connecting seat and a controller, wherein the ball screw mechanism consists of a lead screw, a nut, bearings, bearing seats, a motor and a shaft coupling; the linear guide rail mechanism consists of a guide rail, slide blocks and a sliding plate; one end of the nut sliding plate connecting seat is connected with the nut, and the other end of the nut sliding plate connecting seat is connected with the sliding plate; the controller is electrically connected with the motor. The household ventilating and rain-shielding device is characterized by further comprising a ventilating and rain-shielding plate which is fixed on the sliding plate through a bolt. The household ventilating and rain-shielding device has the advantages that (1) the device is in an open state in the daytime, so that air in southern and northern bedrooms forms convection current, and the aim of natural ventilation is fulfilled; on a cloudy and rainy day, the ventilating and rain-shielding device is closed automatically, so that rain water is prevented from flowing into rooms; (2) bedroom lighting is not influenced; (3) the actions of the device are controlled automatically through the controller.

An intelligent identification and early warning method for abnormal events in the open scene of the electric power field based on edge computing. This invention compresses the improved SSD target detection model and transplants it to the mobile terminal to give full play to the advantages of edge computing. Through experiments, this invention converts the Android end as an optimal solution; the present invention fuses the Conv4_x feature layer and the Conv5_x feature layer in the VGG16 network, and then directly acts on the feature layer after fusion to the final prediction layer, so as to improve the accuracy of small target detection; at the same time , the present invention summarizes a variety of basic weather conditions: sunny, cloudy, rainy, foggy, etc., and uses image enhancement technology to increase training data in different scenarios, thereby improving the generalization ability of the model.

The invention provides a ship target detection method for joint training of deep learning features and visual features, comprising the following steps: sample data collection, CNN feature extraction, traditional invariant moment feature and LOMO feature extraction, feature dimensionality reduction, and feature fusion network FCNN construction , and finally use the sample data to train the network, and use the test data to test the model. Compared with the prior art, the visual feature extraction process of the present invention comprehensively considers the characteristics of ship shape, color and texture, making the detection process interpretable, and can standardize the CNN back propagation process to learn other features than traditional features. This method is fast, efficient, and highly accurate. It still has good detection results for complex scenes such as clouds, cloudy days, and rain, and has high robustness. Features that are complementary to traditional features can be extracted, and the speed is extremely fast, which can achieve the effect of real-time monitoring.