10results about How to "Avoid unreliable" patented technology

The invention discloses a multi-quality skeleton sequence action recognition method based on course learning, and the method comprises the steps: collecting a large number of videos which comprise human body actions and are provided with action category labels, and dividing the videos into a training set and a test set; extracting coordinate information of human body joint points of each frame of image in the video, and generating a continuous skeleton sequence; performing data enhancement on the skeleton sequence; according to the method, a neural network model for action recognition is constructed, then data of a training set is input into the neural network model in batches, the neural network model is trained until a preset round number is reached, in the process of training the model, prior knowledge and posterior knowledge are dynamically adjusted to evaluate sample difficulty, and in addition, the training efficiency is improved. The concept of learning the center of gravity is introduced, and the learning center of gravity gradually transits from an early simple sample to a difficult sample in different stages. The method provided by the invention can improve the recognition efficiency and the recognition accuracy.

The utility model provides a short-wave broadcast deviation switch with self-damping, and the switch comprises a reduction gearbox which is internally provided with a gear transmission mechanism, and the gear transmission mechanism is used for reducing the speed of inputted rotation power and transmitting the rotation power to an output end; and the self-damping device is arranged in the reduction gearbox and connected with the gear transmission mechanism of the reduction gearbox, and when the gear transmission mechanism of the reduction gearbox rotates, the self-damping device can generate a damping effect so as to slow down the rotating speed of the reduction gearbox. By additionally arranging the self-damping device, the problem that the high-power short-wave broadcast deflection switch is unreliable in the deceleration positioning process is effectively solved. The generator in the self-damping device is matched with the power resistor to work, so that a damping effect can be generated when the reduction gearbox rotates, the rotating speed is reduced, mechanical energy can be converted into electric energy, the electric energy is consumed through the power resistor, and effective utilization of the energy is achieved.

The utility model discloses a medicine identification structure of an automatic pharmacy, which comprises a horizontal frame and a vertical frame, a movable bin is arranged on the horizontal frame, and the movable bin is provided with a medicine feeding position and a medicine taking position; the vertical frame is fixed to the side face of the horizontal frame, a medicine feeding manipulator and a medicine discharging manipulator are arranged on the vertical frame, the medicine feeding manipulator is provided with a medicine feeding position, the medicine discharging manipulator is arranged at a medicine discharging position, a medicine feeding recognition code scanning head and a medicine discharging recognition code scanning head are arranged on the vertical frame, the medicine feeding recognition code scanning head is located at the medicine feeding position, and the medicine discharging recognition code scanning head is located at the medicine discharging position. The medicine discharging identification code scanning head is located at the medicine discharging position. Different manipulators act according to identification information of different positions, the problem that a single manipulator is unreliable is solved, and the medicine feeding efficiency of an automatic pharmacy is improved.

This invention provides a method, apparatus, electronic device, and storage medium for detecting vehicle parking status, relating to the field of vehicle detection. The method includes: acquiring a parking space image and a pre-marked parking space area within the parking space image in real time, and determining a target detection mode based on the orientation of a specified parking space frame line within the parking space area; the detection modes include a side-view detection mode and a same-side detection mode; when it is determined that a vehicle detected from the parking space image is stationary, extracting the detection frame corresponding to the vehicle, and sequentially determining the positions of multiple specified chassis key points on the vehicle in the parking space image; when the target detection mode is a side-view detection mode, determining the target parking space area and parking status of the vehicle using the chassis key points and each parking space area; when the target detection mode is a same-side detection mode, determining the target parking space area and parking status of the vehicle using the chassis key points, the detection frame, and the parking space area; and adaptively detecting the vehicle's parking status based on the parking space orientation.

The invention discloses a multi-source information high-precision positioning method and system for an unmanned forklift, and the method comprises the following steps: configuring a millimeter-wave radar, a laser radar and an inertial measurement unit for a dual-radar positioning system, designing a random matching algorithm to calibrate the timestamps of the millimeter-wave radar and the laser radar, and obtaining a main image and an auxiliary image; designing a key frame extraction algorithm to extract key frames of the main image to obtain a primary main image; designing an information matching algorithm, and calculating the positioning of the unmanned forklift by using a dynamic sliding window; a position diagnosis algorithm is designed, positioning deviation is calculated, and positioning of the unmanned forklift is corrected; and outputting the positioning of the unmanned forklift of the double-radar positioning system. According to the scheme, by obtaining the multi-source information of the millimeter wave radar, the laser radar and the inertial measurement unit, a multi-source fusion and dynamic correction mechanism adaptive to a metal processing plant is established, and the positioning precision of the unmanned forklift is remarkably improved.

The utility model relates to the technical field of controllers, in particular to a controller convenient for adjusting the position of a display screen, which comprises an outer shell and a circuit board fixed in the outer shell, a group of transverse female headers are transversely arranged at the top end of the circuit board, and a group of vertical female headers are vertically arranged on the circuit board. Display screens can be inserted into the transverse female headers and the vertical female headers; according to the utility model, the transverse female header and the vertical female header are arranged on the circuit board at the same time, and then the display screen is inserted into the transverse female header or the vertical female header according to specific customer requirements, so that controllers in two forms of a transverse screen and a vertical screen are compatible, and more cost is saved.

The utility model discloses a slice clamping structure which comprises a slice frame provided with a slice groove for containing slices. A first elastic piece and a second elastic piece are arranged on the slice frame and are made of metal materials, and the first elastic piece pushes and extrudes the slice placed in the slice groove in the length direction of the slice groove so as to limit the slice in the length direction of the slice groove. The second elastic piece pushes and extrudes the slice placed in the slice groove in the width direction of the slice groove so as to limit the slice in the width direction of the slice groove; one or more of the first elastic piece, the second elastic piece and the slice frame are provided with limiting parts so that the slices can be limited on the outward side of the slice groove in the depth direction. According to the utility model, the slice is reliably clamped on the slice rack, and the assembly positioning precision is high; the first elastic piece and the second elastic piece are made of metal materials, clamping is easy and convenient, and the problem that the reliability of a plastic elastic arm is not high is solved.

The present invention relates to the field of medical information processing technology, and specifically to an electronic information quality control system for intelligent clinical nursing, comprising: a data acquisition module, used to obtain several types of nursing data for each patient. A data analysis module, used to obtain a first category of reference data; obtain a generalization interval of target data and a second category of target data for each patient according to the first category of reference data, and obtain a generalization interval of target data and a second category of target data for each patient again according to the second category of reference data; and abnormal data in the second category. A quality control module, used to judge the quality of target data of newly added patients according to the abnormal data in the second category and the generalization interval of abnormal data. The present invention timely and accurately judges the quality of target data of newly added patients, while further avoiding the influence of noise caused by individual differences among patients, and ultimately improves the quality of electronic information for clinical nursing.

This application discloses a control method, control device, equipment, and storage medium for a vehicle-mounted unmanned aerial vehicle (UAV), belonging to the field of vehicle control technology. The control method for the vehicle-mounted UAV includes: determining first heading data from multiple heading data sets based on a reference priority, wherein the multiple heading data sets have different data sources; acquiring first image data collected by the vehicle-mounted UAV based on the first heading data set at a first moment; if the first image data satisfies a first switching condition, determining second heading data from the multiple heading data sets based on the reference priority and a first confidence level of the multiple heading data sets at the first moment; and controlling the vehicle-mounted UAV based on the second heading data set. Through pre-configured reference priorities and multiple heading data sets, this application can ensure the accuracy and efficiency of the switching process when heading data switching is required, avoiding blind switching and ensuring the continuity and stability of the vehicle-mounted UAV's automatic takeoff, tracking, and return operations.

The invention discloses a bank insurance data analysis method, which relates to the technical field of data processing. The bank insurance data analysis method comprises the following steps: S1, collecting bank insurance data, and performing standardization processing on the bank insurance data to obtain the customer identity authentication rate and data consistency ratio; S2, performing data cleaning and preprocessing on the bank insurance data to obtain the transaction record error rate and data update frequency; S3, establishing a bank insurance database to obtain the customer duplicate item ratio; S4, obtaining the insurance data quality assessment value according to the data quality original score and the first threshold comprehensive analysis; S5, comparing the insurance data quality assessment value with the second threshold to obtain the insurance data quality assessment report, and specifying corresponding optimization processing measures according to the insurance data quality assessment report. The problem of inconsistent data quality caused by the diversification of bank insurance data sources and uneven data quality is solved.