313 results about "Human monitoring" patented technology

A system and method to improve the automatic adaptation of one or more speech models in automatic speech recognition systems. After a dialog begins, for example, the dialog asks the customer to provide spoken input and it is recorded. If the speech recognizer determines it may not have correctly transcribed the verbal response, i.e., voice input, the invention uses monitoring and if necessary, intervention to guarantee that the next transcription of the verbal response is correct. The dialog asks the customer to repeat his verbal response, which is recorded and a transcription of the input is sent to a human monitor, i.e., agent or operator. If the transcription of the spoken input is correct, the human does not intervene and the transcription remains unmodified. If the transcription of the verbal response is incorrect, the human intervenes and the transcription of the misrecognized word is corrected. In both cases, the dialog asks the customer to confirm the unmodified and corrected transcription. If the customer confirms the unmodified or newly corrected transcription, the dialog continues and the customer does not hang up in frustration because most times only one misrecognition occurred. Finally, the invention uses the first and second customer recording of the misrecognized word or utterance along with the corrected or unmodified transcription to automatically adapt one or more speech models, which improves the performance of the speech recognition system.

Systems and methods for integrated media monitoring are disclosed. The present invention enables users to analyze how a product or service is being advertised or otherwise conveyed to the general public. Via strategically placed servers, the present invention captures multiple types and sources of media for storage and analysis. Analysis includes both closed captioning analysis and human monitoring. Media search parameters are received over a network and a near real-time hit list of occurrences of the parameters are produced and presented to a requesting user. Options for previewing and purchasing matching media segments are presented, along with corresponding reports and coverage analyses. Reports indicate the effectiveness of advertising, the tonality of editorials, and other information useful to a user looking to understand how a product or service is being conveyed to the public via the media.

The invention intends to disclose a method and a system for monitoring and managing videos on the basis of structured description to analyze and understand video images and generate structured description data having correspondence with video data, so that a user can achieve the informationize application in searching, browsing and retrieving video image data through the operation on the video description data. The invention has the advantages of stable and reliable operating performance and wider applicable range, solves the problem of the analysis and management of mass video data, and achieves the purposes of reducing the cost for artificial monitoring and improving the technical levels of the intelligentization and informationization of the existing video monitoring system.

A method for integrated media monitoring is disclosed, wherein multiple forms of media are monitored and searched according to user defined criteria. The method may be used by a business to understand how a product or service is being received by the general public. Monitoring includes analysis of closed captioning data and human monitoring so as to provide a business with a full understanding of advertising and editorial effectiveness. A user provides media search parameters via a network, and a near real-time hit list is produced and presented to the requesting user. Options for previewing and purchasing matching media segments are presented, along with corresponding reports and coverage analyses. Previewing can occur via a streamed video format, whereas purchasing allows for high quality video download. Reports include information about how the product was conveyed, audience watching, and value to the business. Reports can be created by the system or by the user, formatted for presentation, purchased, and downloaded.

The security detection system includes a detection unit capable of detecting a condition, said unit in communication with a remote central server. Where the detection unit is connected to the server via a dial-up modem, the detection system may be connected to a telephone seizure unit. Upon occurrence and detection of a condition, the detection unit prevents a telephone from communicating with a main telephone line, while still allowing said detection unit to communicate electronic information to the server, even upon lifting of the telephone receiver from the base. Once electronic information is sent to the server, a programmed database of the server may send the information according to a pre-determined designation. The server monitors whether a designated party has responded to the information, and if not, may send the information to a human monitoring station. The designated party may request and receive additional information from the detection unit, via the server.

The invention discloses an intelligent processing and search method for social security video monitoring images. Intelligent processing and search of the images are implemented by adopting a social security video monitoring system; and the social security video monitoring system comprises a camera, a video distributor, a video matrix, a display screen, a hard disk video recorder, an intelligent video processor, an alarm device, a network switch, a database server and search workstations. The intelligent video processor are used for performing digitization, coding, pre-processing, target identification, feature extraction and abnormal behavior analysis on the monitored videos, and functions of controlling the alarm device and the video matrix and the like are realized, so that human monitoring is not required, and time and labor are saved; and the search workstations for automatic search in the database server are adopted, so the method is easy to operate, mass research is not required, the research time is saved, and the research result is displayed clearly.

Determination of hydraulic properties such as porosity and permeability of soil is of paramount importance in hydrology and civil engineering. In order to achieve greater accuracy in determination of permeability of soil using falling head permeameter, the two important known constraints of human monitoring error in noting the falling water level between two selected levels and elapse time between these two levels had overcome through electronically sensing the levels between two selected points and activating the timer clock automatically by the pulses coming from senor. The precision in measurement of time lapse in 1 / 100th a second enables greater accuracy in estimation of permeability. Provision of perforated Teflon disc above and below the soil core facilitates in application of water uniformly over the entire surface area of soil core at top and similar way permeated water leaving the soil core uniformly without any obstruction. The use of carbon steel seamless tube while collecting soil core facilitated in undisturbed soil core recovery from desired depth section. The permeability test was conducted for various sorted sands of different size ranges and each sample was subjected to repetitive tests and elapsed time for each test was recorded from timer unit. Coefficient of Permeability was calculated for each test. The lab experiment conducted for sorted and unsorted sediments has yielded a consistent performance of Electronic level sensor and timer based falling head soil permeameter.

The invention aims to disclose a device and a method for realizing video structural description by using DSP and FPGA. The device and the method realize structural description of input video by using the DSP and the FPGA to organically combine an intelligent image processing technique of the video structural description and a DSP and FPGA-based embedded system, analyze and comprehend video images, generate structural description data, effectively monitor the video and manage the video data, and realize information-based application such as query, browse and retrieval of the video image data and the like; and the device has stable and reliable working performance and wide application range, and can greatly improve the intelligence of the conventional video monitoring system, reduce the manual monitoring cost, realize video monitoring information-based management and fulfill the purpose of the invention.

The invention relates to a sleeping cabin capable of monitoring physical signs of human bodies in real time. The sleeping cabin comprises a cabin body, a human body monitoring system, a control system, a server and a user terminal. The control system comprises a calculation module, a display module and an alarm module. The human body monitoring system collects physical sign parameters of a to-be-monitored human body in real time and sends the parameters to the calculation module. The calculation module sends the received parameters to the server. The server analyzes and processes the received parameters, judges whether a grade index of the current health state of the to-be-monitored human body is normal, abnormal or critical and returns a result to the calculation module. The calculation module sends information returned by the server to the display module for display, and triggers the alarm module to give an alarm when the grade index is critical. Compared with the prior art, the sleeping cabin has the advantages that real-time monitoring and early warning effects can be achieved, the sleeping cabin has a great significance in prevention and treatment of sporadic diseases and paroxysmal diseases, and the problems of distribution imbalance of medical resources and poor medical conditions of some regions in China are solved to some extent.

The invention relates to an electric vehicle identification method, device and system. The method comprises: obtaining image information in a preset range of an elevator; identifying whether an electric vehicle exists in the image information; if the electric vehicle exists in the image information, counting the first recognition frequency of the electric vehicle and the second recognition frequency of the electric vehicle which does not exist in preset time; judging whether the first recognition frequency is greater than the second recognition frequency; and if yes, controlling the elevator to execute an electric vehicle taking forbidding instruction. Using the technical scheme, the image information in the preset range of the elevator can be automatically identified, whether the electricvehicle exists in the elevator or not is automatically judged, so that manpower resources are saved. Compared with manual supervision, the scheme for automatically identifying the electric vehicle avoids possible situations such as distracting or carelessness of manual monitoring, reduces monitoring omissions, can reduce potential safety hazards of residents, and guarantees personal safety of theresidents.

The present invention enables service agents in a call center to monitor portions of calls that are especially error-prone (e.g., automated speech recognition of an address, etc.), while portions of calls that are not error-prone occur without any human monitoring. An interactive voice response (IVR) system script (e.g., a VXML script, etc.) is divided into a plurality of independent dialog sequences, each of which is assigned a human-monitoring requirement that indicates whether (and optionally, to what degree) the dialog sequence requires monitoring by a service agent. In the first illustrative embodiment, a partial ordering enables some dialog sequences to appear in a different order in that of the IVR script when it is advantageous to do so. In the second illustrative embodiment, each dialog sequence that requires human monitoring is recorded and packaged into an agent-review task that is subsequently sent to an available service agent for review.

The invention discloses a vehicle illegal parking remote monitoring system. The vehicle illegal parking remote monitoring system is capable of automatically achieving monitoring of illegally-parking vehicles without artificial monitoring, and is high in efficiency. When a vehicle enters a buffer area of a certain illegal parking area, the vehicle begins to be communicated with a server. At the moment, the server can prompt the vehicle. If the vehicle stops in the illegal parking area, the server can automatically warn the vehicle and records in a database. Consequently, the vehicle illegal parking remote monitoring system achieves monitoring to illegal parking behaviors of the vehicles.

The invention discloses a fault self-recovery method for industrial control application program based on heartbeat detection, is applied to industrial control and Internet of Things remote monitoringsystems, and is characterized in that a running state of an application program in an industrial control machine is monitored in real time by monitoring a running situation of an on-site computer ( the industrial control machine), once system halted or abnormal breakdown appears on the industrial control machine or the industrial control application program, the reset of the industrial control machine or restart operation is achieved by a remote monitoring software and embedded equipment, so that the automatic monitoring and automatic recovery of the industrial control machine are achieved, accidents caused by manpower waste and slack of artificial monitoring are reduced so as to guarantee the normal and stable work of the industrial control system.

The invention discloses a dead fish monitoring device and a dead fish monitoring method applied to aquaculture. The device comprises an orbit, a trolley, a PTZ camera and a wireless camera, wherein the PTZ camera is wirelessly connected with the wireless camera and a computer, the computer is connected with an alarm device, an accumulator is installed in the trolley, and a solar cell panel is installed on the top of a support pole which can move up and down and is powered by a cable. Compared with manual monitoring, the monitoring device and the monitoring method provided by the invention can automatically detect dead fishes floating on a water surface, dead fish detection and monitoring efficiencies are improved, and the alarm device connected with the computer can indicate an area where the dead fishes are located when the dead fishes are detected.

The invention discloses a dangerous area human behavior identification and monitoring system based on position and multidimensional information integration. The system comprises an intelligent terminal, wireless base stations, switches and monitoring hosts, wherein the monitoring hosts are connected with the wireless base stations through the switches, the wireless base stations are connected with the intelligent terminal, and a wireless sensing network is formed through the wireless base stations, the intelligent terminal, and the like. The system is advantaged in that multidimensional information sensing capability is realized, multidimensional information integration and behavior identification capability are realized, symmetrical double side bidirectional distance measurement is employed, human or equipment positions are determined in real time, positioning precision is high, anti-interference capability is strong, and human monitoring is carried out in real time through information interaction.

The invention relates to an automatic control method of a mill during the production process of cement and a device thereof. The method is that: a program controller is adopted to control a plurality of links of material powder preparation; a milling sound monitoring instrument is used for monitoring the milling sound change during operation, a dust collector fan inverter is used for controlling the rotating speed, instrument used for detecting the wind pressure in the mill outputs wind pressure signals to the program controller, a controller is used for controlling fan frequency conversion, an inverter is used by a powder-choosing machine for controlling the rotating speed of a rotor of the powder-choosing machine and the program controller adjusts the rotor frequency conversion of the powder-choosing machine according to the input data; a hot blast stove is arranged in front of the mill to raise temperature in the mill and cause water content of the materials to be evaporated in the mill, a thermocouple is applied at the hot wind inlet in front of the mill to monitoring the temperature of the hot wind entering the mill and leading the temperature to rise to a standard that can evaporate water content of the materials in the mill. The invention has the advantages of simple structure, reliable operation, high testing precision, the running of the mill being timely and accurately monitored by operators in a control chamber, thus realizing manual monitoring, reducing labor intensity of workers, increasing output and reducing power consumption.

The invention discloses a prison abnormal condition monitoring method and monitoring system based on depth learning. Generating a sequence feature vector for the audio information based on the secondsub-network; Based on the third sub-network, human behavior is recognized according to the time sequence information of human key points and the sequence eigenvector. Based on the recognized human behavior, abnormal alarm is output when the behavior is abnormal. The prison abnormal condition monitoring method based on depth learning of the invention comprises the following steps of: forming a whole network by three interrelated sub-networks, The training and in-depth learning of the three sub-networks can improve the training speed and recognition accuracy, identify potential abnormal events faster, reduce the cost of human monitoring, eliminate the security risks, and lay a quantitative foundation for the implementation of artificial intelligence to assist the supervision of penalty change and the necessity of detention review and evaluation.

The invention discloses a vehicle tracking and detecting method based on parking lot environment video monitoring and relates to the field of pattern identification, detection and tracking of a moving object and the problem of video monitoring. Through the method of the invention, the monitoring problems of the parking lot environment are obtained more accurately, and compared with the conventional human monitoring method, the method of the invention has the feature of real-time non-interrupted monitoring; and by introducing and using the algorithm of pattern identification, and through the modes of vehicle object detection and correlation analysis of each frame image within time period and the like, the object state of a vehicle is obtained, furthermore, the environment state of a parking lot is obtained, and thus the problem of intelligent monitoring of the parking lot environment is solved.

The invention discloses a bullet cabinet with a remote monitor function, which comprises a cabinet body and a cabinet door, and further comprises a management system used for realizing remote monitoring, wherein the management system comprises a central processing module, an identity identification module, a saving module, a network communication module and a telephone communication module; the central processing module is in two-way connection with the identity identification module, the saving module, the network communication module and the telephone communication module; the identity identification module is used for identifying the identity of an operator; the network communication module is used for remote communication with the client of a high level administrator through a network; the telephone communication module for the remote communication is used for remote communication with a telephone of the high level administrator. Compared with the prior art, the bullet cabinet provided by the invention can further realize remote monitoring for the high level administrator through the network and the telephone on the basis that the manual monitoring and automatic identity identification are combined, so that the safety management of bullet using is greatly completed.

The invention discloses an abnormal behavior detection method, system and device based on an artificial intelligent video. The method comprises steps that S1, video data is acquired; and S2, persons in the acquired video data are detected and tracked, and alarm instructions are sent to alarm equipment when abnormal behaviors of persons are detected. The method is advantaged in that compared with traditional artificial monitoring, false detection and leakage detection probability can be effectively reduced, manpower cost is substantially reduced, real-time monitoring on suspected targets and timely alarm are realized, and social order stability is guaranteed.

The invention discloses a piano practice error correction method based on image recognition processing and relates to the technical field of intelligent teaching. The method comprises the following steps of: step 1, intercepting and shooting images of each frame of finger joints and piano keys from piano teaching videos, and preprocessing each frame of images; step 2, identifying and calculating by using the preprocessed images, outputting piano keys and human body finger joints and position coordinates of the piano keys and the human body finger joints contained in the images, and sequentially connecting the human body finger joints and the piano keys to form a standard finger model diagram and a standard key sequence model diagram. According to the method, the piano practices and the teaching videos are compared and analyzed through image recognition, the practice skill, the key position and the sequence are corrected automatically, errors of each step of a trainer can be found out accurately, guidance reference is provided, the intelligent teaching without human monitoring is achieved, and the difficulty of piano practice can be reduced.

Systems and methods for cleaning and monitoring a portable toilet are provided. The system's components include a chemical detector, a concentration sensor, a cleaning device, a container, a rotatable spray nozzle, a locator, an automated door locking device, a processor, and memory. The memory causes the processor to execute a method which involves detecting and sensing urine and feces concentration levels, and cleaning the areas of highest concentration with chemical neutralizers dispersed by the spray nozzle. Data regarding the sanitary condition and GPS location of the toilet may be sent to servers for remote human monitoring. The door locking device may be triggered if the toilet is found to have harmful levels of chemicals associated with urine and fecal matter.

The invention discloses an oil circulation air-cooling control system of a transformer of a submerged arc furnace. The oil circulation air-cooling control system of the transformer of the submerged arc furnace comprises an air cooling system, a control device, a temperature sensor, an actuating module, an LCD (liquid crystal display) and an operating module, wherein the air-cooling system comprises a circulating oil system and an air cooler and is connected with the actuating module; the control device comprises a programmable logic controller (PLC), a wireless communication module and a remote controller; the remote controller is wirelessly communicated with the programmable logic controller (PLC) remotely through the wireless communication module; the temperature sensor detects the temperature of oil on the top of an oil tank of the transformer in real time, and transmits acquired oil temperature signals to the programmable logic controller (PLC); and the LCD and the operating module are connected with the programmable logic controller (PLC). The air-cooling control system has a remote control function, an intelligent monitoring function and a manual monitoring function, and can accurately and efficiently control an air-cooling system of the transformer in real time from multiple functions; and long-term normal running of equipment is guaranteed.

The invention discloses an urban inland inundation monitoring method based on a deep learning technology. The urban inland inundation monitoring method comprises the following steps: step 1, making aninitial ponding image data set; 2, preprocessing the initial ponding image data set; step 3, performing training based on a Mask RCNN instance segmentation algorithm, and obtaining an optimal pondingmodel water.h5; 4, calibrating the camera by adopting a method based on a single-plane checkerboard, and carrying out distortion correction on the image; 5, performing perspective transformation processing; and step 6, on the basis of a water.h5 model, taking a result image subjected to perspective transformation in the step 5 as model input, and calculating and obtaining a real ponding area range. According to the method, the limitation of a traditional manual monitoring method is broken through, the real-time performance and the accuracy of ponding information acquisition in the waterlogging monitoring process are guaranteed, and powerful technical support is provided for efficiently, safely and quickly carrying out urban waterlogging monitoring work.

The invention relates to a self-adaptive intelligent concrete distributing method and system which are used for controlling pouring operation of a concrete distributor. The distributing method comprises the following steps of carrying out distributing position planning by using a BIM technology and obtaining a corresponding distributing path; and when the concrete distributor conducts concrete pouring, controlling a distributing opening of the concrete distributor to move along the distributing path, and accordingly completing the pouring operation. According to the distributing method, the distributing position is planned by combining the BIM technology, the distributing path is obtained, the distributing opening of the concrete distributing machine is controlled to move through the distributing path, distributing automation is achieved, manual assistance and manual monitoring feedback are omitted, the intelligent level is improved, the labor intensity is relieved, and the problems that time and labor are wasted, labor cost is high, and monitoring feedback is not timely are solved.

The invention relates to the technical field of high voltage power transmisssion operation and maintenance and particularly relates to a recognition and early warning system and method for preventing the large mechanical damage of a transmission line. A monitoring device in a tower carries out video collection on a protection area and transmits a video to a remote monitoring background terminal through a wireless communication network. The remote monitoring background terminal processes a video image, judges whether the protection area has large mechanical invasion and sends screenshot information to an image storage cloud server regularly. A user can obtain an image from the image storage cloud server and carry out further judgment. According to the video image processing, large machinery is judged, the video image processing is more targeted, intelligent monitoring and human monitoring are combined, and the damage of power equipment caused by a system judgment error can be avoided.

The invention discloses an infusion dripping speed monitoring device. A manual monitoring mode is adopted for current infusion dripping speed monitoring management. The infusion speed monitoring device comprises a shell, and a lithium battery and a control circuit are arranged in the shell. The control circuit comprises a control module, a battery charging module, a power supply control module, a human-computer interaction module and a signal acquisition module. The shell comprises a front shell and a rear shell; a liquid crystal display hole and indicating lamp holes are reserved in the face of the front shell; the rear shell comprises an integrally formed main shell body, a positioning block and two monitoring element mounting blocks; a data interface hole and key holes are reserved on the side wall of the main shell body, and buzzer holes are reserved on the bottom surface of the main shell body; the positioning block and the two monitoring element mounting blocks are all arranged on the back of the main shell body; a monitoring element mounting groove is reserved in each monitoring element mounting block; a positioning hole is reserved in the positioning block; part of the side wall of the positioning hole penetrates one end of the positioning block through a line through groove. Dripping speed of an infusion tube can be monitored accurately, and the device can be operated with one hand, and is simple and convenient.

The invention provides a warehouse monitoring management device and relates to warehouse management field. The warehouse monitoring management device receives the location measurement information sent by the location measurement information collector, and determines whether the monitored goods move in accordance with the location measurement information received after the rest time of monitored goods associated with the location measurement information collector exceeds the pre-set time; If the monitored goods move, the device sends an alarm instruction to an alarm. The warehouse monitoring management device can monitor the status of the monitored goods in a real-time manner and notify the alarm to alarm in case of any abnormality, with no human monitoring involved, which greatly reduces the labor cost, with no damage risk for monitored goods and high reliability.

The present invention enables service agents in a call center to monitor portions of calls that are especially error-prone (e.g., automated speech recognition of an address, etc.), while portions of calls that are not error-prone occur without any human monitoring. An interactive voice response (IVR) system script (e.g., a VXML script, etc.) is divided into a plurality of independent dialog sequences, each of which is assigned a human-monitoring requirement that indicates whether (and optionally, to what degree) the dialog sequence requires monitoring by a service agent. In addition, instances of the dialog sequences in calls are prioritized based on an indication of caller intelligibility during (i) prior dialog sequences in the current call, and (ii) prior calls involving the same caller and / or the same contact identifier (e.g., telephone number, email address, Internet Protocol address, etc.).