38results about How to "Automatic judgment" patented technology

The invention discloses an action video extraction and classification method based on moving object detection, and belongs to the field of big data artificial intelligence deep learning. The method comprises the following steps of S1, collecting motion videos of various scenes and different qualities, establishing a motion video database, classifying and marking human body actions in the motion video database, and establishing a motion action library; S2, establishing a detection model of the moving object based on the video image by utilizing a deep learning technology; S3, training the moving target detection model by using the moving action library; and S4, taking the video shot by the user in real time as the input of the moving target detection model, judging the human body action category, and judging whether the action video needs to be extracted and stored in combination with the user demand. The method provided by the invention can automatically, accurately and quickly complete judgment of the motion category and the motion starting moment, and can adapt to video extraction and classification tasks in different scenes in real time.

The invention discloses a method for automatically judging whether tourists fall behind, which adopts the technical scheme for effectively monitoring the fallen behind tourists in group tour: calculating the Euclidean distance between each tourist and other n-1 tourists to obtain a distance sum; analyzing the distance sum, screening out special tourists, calculating average value of the distance sum of the rest of the tourists, and calculating the relative continuous distance between two tourists according to the average value so as to generate a relative continuous area; acquiring intersection of all relative continuous areas and performing calculation to obtain a main branch group and an auxiliary branch group; and judging the containing relation of the special tourists and the branch groups the special tourists belong to, and finally determining the fallen behind tourists according to fallen behind tourist judging conditions. The method can judge the fallen behind tourists timely, accurately and automatically, and can estimate the current dispersion degree of the group according to the relative continuous distance between two tourists so as to avoid and reduce phenomena of tourists falling behind.

The invention discloses an automatic real-time solar radio burst detection method based on image processing. According to the method, the detection of solar radio burst is regarded as moving object detection for a solar radio spectrum image, the idea of a Gaussian mixture model is adopted to detect the moving region of the solar radio spectrum, accurate quantization is performed on the detected burst region, the burst strength, burst bandwidth and duration are extracted, and two parts, namely the automatic detection of solar radio burst and the extrication of parameter values of the solar radio burst are included. According to the invention, the parameter values of the solar radio burst are extracted automatically, the detection work of solar radio burst, which is completed manually and wholly, is completed automatically through an image processing method, so that efficiency and accuracy of parameter extraction are improved; the solar radio burst detected in real time and extracted burst parameters are provided as data for early warning of severe space weather in real time; before the severe space weather, the modern equipment system is adjusted, so that the influence brought by severe space weather is reduced as far as possible.

The invention provides a method for preventing server hard disk failures caused by abnormal environmental temperature and humidity and relates to the technical field of computer storage. The method comprises (1) the requirements of hard disk parts for the environmental temperature and humidity, (2) a method for collecting environmental information regularly, and (3) a script program for comparing values automatically. According to the method, the environmental temperature and humidity information is collected regularly through a scheduled task function in a Linux operating system and stored for timed comparison, comparison is performed with the required value of each hard disk part of a server for the environmental temperature and humidity as a reference value, and an anomaly in a use environment can be found timely.

The invention discloses a spectrum calibration method, a spectrum calibration device and spectrum calibration equipment of an echelle spectrometer, and a storage medium. The spectrum calibration method comprises the steps of: calculating coordinate deviation values of light spot positions of a mercury lamp of a current frame at different wavelengths by utilizing an image processing algorithm; andgenerating an instruction that a rotation angle of a detector assembling and adjusting position does not meet requirements if the coordinate deviation values are not within a deviation allowable range, wherein the deviation allowable range is determined based on a coordinate deviation standard curve and a threshold value, the coordinate deviation standard curve is obtained by fitting the coordinate deviation values of light spot positions with different wavelengths and corresponding position deviation values, and the position deviation values are deviation values of a theoretical position of acharacteristic spectrum design model and an actual position of a detector light spot at the optimal matching position. According to the spectrum calibration method, the spectrum calibration device and the spectrum calibration equipment, the problem that the spectral calibration precision is reduced due to the rotation of the light spot positions relative to an optical system in the prior art is solved, the installation and debugging precision of the echelle spectrometer detector is ensured, and the spectral calibration precision of the echelle spectrometer is effectively improved.

The invention provides a terminal data processing system. The system comprises a server, a database and a storage medium which stores computer programs, and the server is in communication connection with the database; and terminal data is stored in the database, the terminal data comprises a terminal ID, a historical grid of the terminal, uploading time t1 of the historical grid, a WIFI-ID connected with the terminal and uploading time t2 of the WIFI-ID, and the server executes the computer programs to complete corresponding steps and achieve corresponding functions.

The invention discloses a method and system for determining production progress and order inventory, and an edge server, and the method comprises the steps: collecting a video frame of a production link operation region in a production line; performing image recognition on the acquired video frame, and determining the brightness of an operation area in the video frame; and when the brightness of the operation area in the video frame is smaller than a preset brightness threshold value, determining that the production link operation is finished, and obtaining the finishing time of the productionlink operation. By the adoption of the scheme, non-invasive, automatic and accurate judgment of the production progress can be achieved.

The invention discloses an activation data exception detection and analysis method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the stepsof determining a current feature vector according to current activation data and historical activation data of a channel; determining the similarity between the current feature vector and a referencefeature vector in a feature vector library; and determining whether the current activation data is exceptional or not according to the similarity. According to the embodiment, the technical defects that in the prior art, when whether the activation data of the channel is exceptional or not is detected, the threshold value is difficult to determine, and the activation data cannot be comprehensivelydetected are overcome; therefore, the beneficial effects that whether the activation data of the current channel is exceptional or not is accurately, comprehensively and automatically judged, subsequent analysis of reasons for exception of the activation data is facilitated, and manual intervention is not needed are achieved.

The invention discloses a dislocation detecting and self-repairing control method for temperature detectors of a dual-compressor air conditioning system. The method comprises the steps that a, a control circuit is set up; b, a unit is started to operate for a minutes, one compressor (12) is started, the other compressor (12) is shut down, judgment is carried out, if one of the following two conditions is met, inserting dislocation of the temperature detectors is judged, according to the condition 1, detection is carried out once every b minutes, the exhaust temperature of the compressor (12) which is not started is increased by m DEG C or above, meanwhile, the exhaust temperature of the started compressor (12) is decreased by n DEG C or above, and according to the condition 2, the shutdown time of the compressors (12) exceeds c minutes, and the exhaust temperature of the compressors (12) is larger than or equal to z DEG C; and c, a master controller (1) turns on two normally open switches and turns off two normally closed switches. According to the method, whether inserting dislocation of the temperature detectors happens or not can be detected fast and accurately, and the dislocated temperature detectors can be automatically corrected to the correct compressors (12).

The invention discloses an automatic classifying and stacking device for stationery case production, which comprises a box body, an upper adjusting disc and a lower adjusting disc are fixedly arranged on the upper side and the lower side in a working bin respectively, a plurality of first adjusting magnetic blocks are fixedly arranged on the upper adjusting disc, and a plurality of second adjusting magnetic blocks are fixedly arranged in the lower adjusting disc respectively. A plurality of adjusting air pipes are vertically arranged between the second adjusting magnetic blocks in the lower adjusting disc, a moving posture camera is fixedly arranged at the top in the classification conveying box, a plurality of third conveying magnetic blocks are fixedly arranged on the upper surface and the lower surface in the classification conveying box correspondingly, and a sealing plate slidably sleeves a positioning column; and a plurality of placing frames are fixedly arranged on the side, located in the classification conveying box, of the sealing plate. According to the writing case, the controller and the current regulator are used for controlling the on-off of the adjusting magnetic block I, the adjusting magnetic block II, the transmission magnetic block I and the transmission magnetic block II and the intensity of magnetic force, so that the writing case is suspended, moved and rotated.

The invention provides a terminal data processing system, which comprises a server, a database and a storage medium storing computer programs, and is characterized in that the server is in communication connection with the database; the system is characterized in that terminal data are stored in the database, the terminal data comprise terminal IDs, historical grids of terminals, uploading time t1of the historical grids, WIFI- IDs connected with the terminals and uploading time t2 of the WIFI -IDs, and the server executes the computer programs to complete corresponding steps and achieve corresponding functions.

The invention provides a method, device and system for detecting skew of an unmanned aerial vehicle. The method comprises the steps that after the unmanned aerial vehicle is powered on, controlling the unmanned aerial vehicle to hover at a fixed point in a position mode; after the unmanned aerial vehicle successfully hovers at a fixed point, controlling the unmanned aerial vehicle to be switched from a position mode to an attitude mode; when the unmanned aerial vehicle is kept for a preset first duration in the attitude mode, obtaining the current horizontal speed of the unmanned aerial vehicle; judging whether the current horizontal speed of the unmanned aerial vehicle is greater than a preset horizontal speed threshold value or not; if yes, determining that the unmanned aerial vehicle has the skew problem, otherwise, determining that the unmanned aerial vehicle does not have the skew problem, or sending the current horizontal speed of the unmanned aerial vehicle to a ground control station of the unmanned aerial vehicle, and judging whether the unmanned aerial vehicle has the skew problem or not by the ground control station according to the current horizontal speed of the unmanned aerial vehicle. Whether the unmanned aerial vehicle is inclined or not can be automatically and accurately detected.

The invention relates to a script information processing method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a reference object identification in an object introduction text corresponding to each session in a script file, and obtaining a reference object identification set; in the plot text corresponding to each session in the script file, obtaining a to-be-analyzed object identifier with a dialogue attribute to obtain a to-be-analyzed object identifier set; calculating a matching result between the reference object identifier set and theto-be-analyzed object identifier set in each session; and determining the accuracy information of the object introduction text in the script file according to the matching results corresponding to all the sessions in the script file. According to the embodiment of the invention, the matching result can be obtained through matching calculation of the reference object identifier set and the to-be-analyzed object identifier set, and the accuracy information of the object introduction text in the script file is automatically determined based on the matching result, so that automatic actual outgoing person judgment can be realized, manual check is not needed, the efficiency is improved, and the labor cost is reduced.

The invention provides a terminal data processing system, which comprises a server, a database and a storage medium storing computer programs, and is characterized in that the server is in communication connection with the database; the system is characterized in that terminal data are stored in the database, the terminal data comprise terminal IDs, historical grids of terminals, uploading time t1of the historical grids, WIFI- IDs connected with the terminals and uploading time t2 of the WIFI- IDs, and the server executes the computer programs to complete corresponding steps and achieve corresponding functions.