14results about How to "Improve determination accuracy" patented technology

This application discloses a method, device, electronic device, and storage medium for obstacle avoidance navigation for the blind, relating to the field of computer technology. The method includes: performing target detection on the current environmental image of a preset environmental area to obtain the obstacle type of each obstacle; determining the obstacle weight factor of the corresponding obstacle based on the obstacle type of each obstacle; and determining the speed sensitivity coefficient of the corresponding obstacle based on the current obstacle data of each obstacle and the historical obstacle data of the corresponding obstacle; determining the noise interference coefficient of the preset environmental area at the current moment based on the current signal strength in the current voice data of the preset environmental area; and determining the obstacle avoidance risk index of the corresponding obstacle based on the noise interference coefficient, the current obstacle data of each obstacle, the obstacle weight factor, and the speed sensitivity coefficient; and performing graded obstacle avoidance navigation for the blind based on the obstacle avoidance risk index of each obstacle, which can effectively help the blind avoid dangers in complex environments.

Embodiments of the present specification provide an object inventory method, an object inventory device, a computer readable storage medium, an electronic device and a computer program product. The method comprises: performing target detection on an image about a target region to obtain a plurality of detection boxes about objects in the target region, and then needing to identify the category of the object corresponding to each detection box. In the embodiments of the present specification, the actual category corresponding to each detection box is determined based on a seed object of a known category. Further, according to the actual category corresponding to each detection box, the inventory data of the objects in the target region is determined, including: the number of categories of the objects contained in the target region, and / or the number of objects belonging to each category in the target region.

The embodiment of the invention provides an image processing method and device, electronic equipment and a storage medium, belongs to the technical field of artificial intelligence, and can be applied to scenes such as finance, smart city and medical treatment. The method comprises the following steps: performing instance segmentation on a to-be-processed image to obtain a first feature map, and determining a weight corresponding to the first feature map; determining a foreground suppression graph corresponding to the to-be-processed image according to the weight and the first feature graph, and performing convolution operation on the foreground suppression graph to obtain a second feature graph; according to the feature vectors of the pixels in the second feature map, performing momentum clustering on the pixels in the second feature map to obtain clustering distribution probability values corresponding to the pixels in the second feature map; extracting a target region in the second feature map according to the clustering distribution probability value; and determining the similarity between the two to-be-processed images according to the target areas of the two to-be-processed images in the image pair. According to the embodiment of the invention, the accuracy of determining the similarity between the images can be improved.

The invention relates to display equipment and an image display method, and relates to the technical field of display equipment. The display apparatus includes a display, a backlight assembly, and a controller. Wherein the backlight assembly comprises a plurality of backlight subareas, and each backlight subarea comprises a plurality of color backlight sources; the controller is respectively coupled with the display and the backlight assembly, and is configured to determine a target color value to be displayed in the display by each pixel in an input image; obtaining a compensation image of the input image and backlight information of each backlight partition corresponding to the input image according to the target color value and the light transmittance information of the display; and controlling the display to display the compensation image, and adjusting the brightness of the color backlight source in each backlight subarea according to the backlight information of each backlight subarea. The image display effect of the display equipment can be improved.

The invention relates to a power switch device loss separation test method and device, computer equipment, a storage medium and a computer program product. The method comprises the steps of determining discharge loss corresponding to a power switch device according to first drain-source voltage data and first drain current data of the power switch device to be tested in a conduction steady state, and determining discharge loss corresponding to the power switch device according to second drain-source voltage data and second drain current data of the power switch device in a turn-off transient state, determining the overlapping loss corresponding to the power switch device, and determining the charging loss corresponding to the power switch device according to the third drain-source voltage data and the third drain current data of the power switch device in the turn-off steady state; determining hysteresis loss corresponding to the power switch device according to the charging loss, the discharging loss and the overlapping loss; and obtaining a loss test result corresponding to the power switch device based on the charging loss, the discharging loss, the overlapping loss and the hysteresis loss. By adopting the method, the loss test precision of the power switch device can be improved.

This disclosure provides an information processing method and apparatus, a communication device, and a storage medium. The information processing method includes: determining the time domain location of a predetermined TDW based on the antenna switching interval and the antenna switching time.

The application relates to a transformer oil detection method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: performing separation treatment on transformer oil to be detected to obtain a to-be-detected gas corresponding to the transformer oil; the transformer oil is obtained by collecting a plurality of oil taking ports of a to-be-detected transformer; an optoacoustic signal corresponding to the to-be-detected gas is acquired; according to the optoacoustic signal, a current acetylene content value corresponding to the to-be-detected gas is determined; a feature vector of the current acetylene content value is extracted; the feature vector is respectively input into a plurality of trained transformer oil detection models to obtain target prediction probabilities of the transformer oil under various preset detection results; from the various preset detection results, a preset detection result with the maximum target prediction probability is screened out as a target detection result corresponding to the transformer oil. By adopting the method, the detection accuracy of the transformer oil can be improved.

The embodiment of the specification discloses a subtitle display method, device and equipment, the method comprises the following steps: a long connection is established with a first device, and the target voice stream to be converted collected by the first device is received through the long connection; text conversion processing is performed on the target voice stream, first text data corresponding to the target voice stream is obtained, and correction processing is performed on the first text data, second text data corresponding to the target voice stream is obtained; based on the first text data and the second text data, it is determined that the subtitle is target text display data corresponding to the target voice stream. Through the above-mentioned subtitle display method, a long connection can be established with the first device, the resource utilization rate is improved, and the text data display efficiency and accuracy of the target voice stream are improved.

An exposure adjustment method and device, a computer readable storage medium and a terminal, the method comprising: extracting a subject saliency map in an original image; determining a target detection region based on the subject saliency map; determining the luminance values of each sub-region of the original image, the subject saliency map, the target detection region and a background region respectively; determining the weight values of each grid region in the original image, the subject saliency map, the target detection region and the background region respectively; performing weighted operation on the luminance difference values of the original image, the subject saliency map, each sub-region of the target detection region and the background region and the weight values of the corresponding grid regions to obtain a to-be-adjusted luminance difference value; and adjusting the exposure of a shooting device of the original image using the to-be-adjusted luminance difference value. The present application can improve the matching degree of the to-be-adjusted luminance difference value and the actual subject condition of the image, and improve the exposure adjustment accuracy.

The invention relates to the technical field of robot navigation, and discloses a robot navigation method and device based on multi-source data fusion, and the method comprises the steps: carrying out the fusion processing of the multi-source data of a robot, obtaining the environment map of the current environment of the robot and the pose of the robot, and obtaining the position of the robot according to the visual image data and the laser point cloud data; obstacle information in the current environment is recognized, the optimal path of the robot from the current position to the target end point position is determined according to the environment map, the robot pose and the obstacle information, a motion control instruction is generated according to the optimal path, and the robot is controlled to run according to the motion control instruction. Therefore, by implementing the method, the determination accuracy of the map and the pose of the robot can be improved through multi-source data fusion, the obstacle recognition accuracy is improved, and the stability, high efficiency and safety of autonomous navigation of the robot are further improved.

The application provides a method and device for determining a cardiac risk parameter, the method comprising: obtaining a plurality of lead electrocardio vector signals of a target entity; converting the plurality of lead electrocardio vector signals into a plurality of time domain waveform graphs; obtaining a waveform difference value between a waveform maximum value and a waveform minimum value of each time domain waveform graph to obtain a plurality of waveform difference values; obtaining a time cost for each time domain waveform graph to reach the waveform maximum value to obtain a plurality of time costs; obtaining a time interval between two adjacent waveform maximum values on each time domain waveform graph to obtain a plurality of time intervals; determining an electrocardiogram risk parameter based on a standard deviation and a maximum value of the plurality of waveform maximum values, a standard deviation and a maximum value of the plurality of waveform difference values, a standard deviation and a mean value of the plurality of time intervals, and a standard deviation and a mean value of the plurality of time costs; and determining the cardiac risk parameter based on the electrocardiogram risk parameter. The application can improve the determination accuracy of the cardiac risk parameter.

The application discloses a kind of based on circuit board's wiring length determination method, it is related to circuit board technical field.According to the current layer of circuit board, the coordinate corresponding to the auxiliary area of the target device connected with the starting point of the current wiring when the current wiring reaches is determined to the wiring length of the current layer, considering the actual coordinates of the vicinity of the target device reached by the starting point of the current wiring, avoid the problem of statistical error caused by artificial statistics to determine the wiring length of the corresponding target device.In the case of via, the starting point of the corresponding wiring of the next layer needs to be tracked by the via tracking, to avoid the omission of artificial statistics.In the case of pin, the fact that the corresponding pin device includes two pins in the conventional artificial statistical process will ignore the statistical error caused by the wiring length between the pins.Considering the wiring length determination of the wiring length mapping relationship between the inner layer and the outer layer of the circuit board, the calculation accuracy of the wiring length is improved.

This invention provides a robot visual servo control method for operating biochemical instruments. The method includes: performing visual detection on the working area of ​​the biochemical instrument to acquire effective real-time images and hand-eye calibration parameters; extracting the target contour features of the target to be operated; generating an initial visual pose and determining the near-field deployment area; deploying a disposable near-field reference body at a non-operational position to obtain the reference body contour features, deployment confirmation results, and effective co-field images; generating local refraction observation data based on the effective co-field images, reference body contour features, disposable near-field reference body specification markings, and hand-eye calibration parameters; solving for the local refraction offset results and constructing a visual correction matrix; generating the actual operating pose, updated visual servo mapping relationship, corrected motion trajectory, current position reference sequence, and execution control queue based on the visual correction matrix; and performing dynamic visual detection to generate a current position detection sequence.

This application relates to the field of fly ash ceramsite preparation technology, specifically to a method for preparing steam-cured ceramsite with high fly ash content. The method includes: weighing raw fly ash and an activator for ball milling; obtaining ultrafine fly ash through sieving after ball milling; mixing and granulating the ultrafine fly ash with a gel material to obtain ceramsite green bodies; curing the ceramsite green bodies in an automated constant temperature and humidity curing chamber; collecting environmental parameters at various monitoring points within the curing chamber at different times, determining the range of influence of these environmental parameters at the current time, using this as the prediction time-domain step size for a model predictive control algorithm, and controlling the environment within the curing chamber using the model predictive control algorithm; and sending the cured ceramsite green bodies into a fully automatic autoclave for further curing, adjusting the proportional parameters of a PID controller to control the pressure within the autoclave, and obtaining steam-cured ceramsite after further curing. This application improves the preparation performance of steam-cured ceramsite.