49results about How to "Solve the problem of misjudgment" patented technology

The invention provides a robot control method, a device and a storage medium. The method comprises the steps of receiving voice messages, and extracting character information in the voice messages; determining whether preset characters exist in the character information; determining the position of a target object according to the voice messages if the preset characters exist in the character information; performing gesture recognition on the target object so as to obtain gesture recognition results; and inquiring an instruction library according to the gesture recognition results so as to determine corresponding control instructions, and controlling a robot to perform the corresponding operation according to the control instructions, wherein the instruction library contains different gesture recognition results which are set in advance and the corresponding control instructions.

The invention is applicable to the technical field of interdisciplinary of deep learning and underground pipe gallery engineering, and relates to an underground drainage pipeline disease segmentationmethod based on a full convolutional neural network, which comprises the following steps: acquiring an underground drainage pipeline disease data set; making a drainage pipeline disease data set; optimizing a semantic segmentation algorithm; making model hyper-parameter adjustment; performing model training; verifying a model; and testing the model. A deep learning algorithm is adopted, an FCN full convolutional neural network is optimized, a semantic segmentation method suitable for complex and similar disease characteristics of the underground drainage pipeline is researched and developed, and real underground drainage pipeline disease detection big data is adopted, so that segmentation of the disease pixel level of the underground drainage pipeline is realized, better robustness and generalization ability are achieved, and the invention is suitable for large-scale popularization and application. Precision and efficiency of underground drainage pipeline disease detection are effectively improved.

The invention discloses a code leakage detection system and method based on a natural language processing technology. The method comprises the steps: taking a code as a stricter and more logical special language, carrying out the analysis of a source code through the natural language processing technology, replacing the manual analysis and analysis process, and extracting the key information whichcan most represent a project; extracting webpage information by using a web crawler technology, and performing content screening extraction on the acquired webpage information by using a machine learning technology; using a strategy of combining a traditional text similarity calculation method and a deep learning algorithm, carrying out similarity comparison on source code and an online suspicious code, displaying suspicious webpages according to the similarity, and providing a basis for final judgment. The problems that when source code is analyzed through a traditional manual means and keyinformation is extracted from the source code, the extraction excessively relies on a human experience, and large-scale code analysis is low in efficiency, time-consuming, financial-consuming and thelike are solved.

The present invention relates to a method of judging the type of digital versatile disc and a device. The method includes the steps: setting relative parameters, and obtaining a wobble signal from the digital versatile disc, setting frequency removing factor, and removing frequency for the wobble signal by the frequency removing factor; then, calculating the frequency of the wobble signal after frequency removing taking the time that the main shaft motor of a disc player drives the digital versatile disc rotate for one circle as the unit time; and last, judging the type of the digital versatile disc based on the frequency of the wobble signal after frequency removing, thereby can avoid the former fault and improving the correct judging probability.

The present application provides a robot control method, device and storage medium, the method includes receiving voice information, extracting character information in the voice information; judging whether there are preset characters in the character information; if yes, determining the target according to the voice information The position of the object; perform gesture recognition on the target object to obtain the gesture recognition result; query the instruction library according to the gesture recognition result to determine the corresponding control instruction, and control the robot to perform the corresponding operation according to the control instruction; where the instruction library contains different gesture recognition presets Results and their corresponding control instructions.

The invention discloses a method and a device for judging the detection result of a treponema pallidum antibody. The method comprises the following steps: performing agglutination reaction on serum or plasma of a to-be-detected patient by the treponema pallidum antibody and performing reaction on the serum or plasma of the to-be-detected patient without the treponema pallidum antibody separately by a treponema pallidum gelatin agglutination test method to acquire respective sample reaction plates; obtaining pictures of a to-be-detected sample and the control sample reaction plates; performing imaging processing on the pictures to acquire a plurality of pixel points and corresponding agglutination color pixel values from the pictures, and converting the pictures into two-dimensional data for storage; comparing parameters, such as shape, peak value, ratio of peak height to peak width, peak valley, difference value between peak height and peak valley and agglutination ring size, of agglutination color gray value curves of a to-be-detected group and a control group with a standard threshold value to judge the detection result. According to the method and the device for judging the detection result of the treponema pallidum antibody, the detection result of the treponema pallidum antibody can be judged more accurately and the detection efficiency is improved.

The invention discloses a repair detection method and device of a display module, equipment and a storage medium. The repair detection method comprises the steps of obtaining a first anode current of a light-emitting element in a bad sub-pixel, and calculating first compensation data according to the first anode current; acquiring a second anode current corresponding to light-emitting elements in at least two pixel groups except the bad sub-pixel, and calculating second compensation data corresponding to the pixel group according to the second anode current corresponding to the pixel group; determining the maximum difference value in the difference values of the second compensation data and the first compensation data; judging whether the maximum difference value is greater than or equal to a preset threshold value or not; if so, prompting that the bad sub-pixel is successfully repaired; and if not, prompting that the repair of the bad sub-pixel fails. The repair detection method is used for detecting whether the display module is successfully repaired or not, the detection mode is more intelligent, and the accuracy of the detection result can be improved.