91results about How to "Detection intelligence" patented technology

The invention discloses an indoor thermal comfort evaluation system, which is characterized in that by taking six parameters, namely the indoor air temperature, the air relative humidity, the air velocity and the black globe temperature collected by a sensor module and the human metabolic rate and the clothing thermal resistance input by a key module as the input signals, and establishing a simulation evaluation model in a singlechip module through a neural network, the PMV-PPD thermal environment comfort index is intelligently obtained. The indoor thermal comfort evaluation system has high integration degree, can objectively and intelligently evaluates the indoor thermal comfort, and effectively avoids the troublesome formula iterative operation.

The invention discloses a bearing defect automatic detection system based on multi-angle light source images. The bearing defect automatic detection system comprises an acquisition unit, a PC end processing unit, a data analysis unit, an intermediate processing unit and a defect detection unit. According to the system, image capturing detection is carried out under the multi-light-source illumination condition, the tiny error can be better detected, the problem of identifying a specific identifier on a bearing is solved, and detection is more intelligent. The system not only comprises softwarein the aspect of bearing image processing, but also independently builds a matched detection processing hardware framework, and simulates the industrial automatic bearing detection process, so that the whole system can smoothly run to achieve the purpose of automatically screening bearings, and meanwhile, the system is more suitable for practical application.

The invention discloses a red light running detecting device and method based on video processing and belongs to the fields of image pattern recognition, video target tracking, intelligent video monitoring and intelligent transportation. The red light running detecting method comprises a license plate locating method based on edge information and color information and a vehicle target tracking method based on feature points of an angular point. According to the red light running detecting method, the most obvious partial characteristics of an image of a vehicle target are used as the basis of vehicle locating, relevance of the characteristics of a rigid body of the moving vehicle of multiple frames is used for tracking the vehicle, and therefore movement information of the vehicle target in a monitoring video can be accurately obtained. The red light running detecting device and method based on video processing can rapidly and accurately detect the position of a license plate of the vehicle in a monitoring video, can accurately obtain the movement information of the vehicle target, can intelligently achieve the function of vehicle running snapshot, can intelligently achieve the function that whether the vehicle runs on a required lane is judged, and is applicable to a red light running electronic police system, a public security access system, an abnormal-vehicle event detection and no-parking automatic fare collection system and the like.

The invention discloses a method, a device and a system for detecting food. The method for detecting the food comprises the following steps: receiving a detection signal transmitted by tableware, wherein the detection signal is generated by detecting the state of the food through a built-in sensor of the tableware; and processing the detection signal to generate a detection result, and displaying the detection result on a terminal screen. According to the method, the device and the system for detecting the food, the detection signal which is transmitted by the tableware is received, the detection signal is processed to generate the detection result, and the detection result is displayed on the terminal screen, so that the state of the food can be rapidly and intelligently detected, the state of the food can be acquired by a user in time, reliable evidence can be provided for the user to judge whether the food is safe to eat, and the diet health of people can be effectively guaranteed.

The invention provides a method and a device for detecting whether a capacitor is correctly mounted through photographing. The method comprises steps: an image of a circuit board of an already-assembled capacitor is acquired through a camera; the acquired image is pre-treated, wherein pretreatment comprises graying, histogram transformation, binarization and expansion corrosion treatment on the image; noisy points with over small areas on the image after pretreatment are removed, the capacitor is found through Hough circle transformation, and invalid circles are eliminated; and whether the capacitor is correctly mounted is judged according to whether a negative electrode marker exists in an effective circle preset position. The method is measurement based on computer vision, is a non-contact measurement technology, particularly a capacity polarity detection method brought forwards based on open source OpenCV, and solves the defects of slow execution speed and poor real-time performance of the traditional detection method.

The invention provides a method and device for detecting a behavior of latently stealing user data, and relates to the technical field of network information security. The method comprises the steps of obtaining a first access log to be detected, wherein the first access log is obtained when a user accesses a website; calculating a first user access behavior feature value according to the first access log; inputting the first user access behavior feature value to a detection model which is built in advance to detect whether the behavior of latently stealing the user data exists in the access log or not. The detection model is used for detecting the behavior of latently stealing the user data, and is obtained through classification training of training sample data. According to the method and device method for detecting the behavior of latently stealing the user data, the detection model for detecting the behavior of latently stealing the user data and model parameters can be obtained by calculating the user access behavior feature value of the access log and conducting training of a classification model, and then the detection model is utilized to precisely, efficiently and intelligently detect the behavior of latently stealing the user data in a new access log.

The invention provides a device for detecting a steel tie rod or an inhaul cable and application. A rod section detection mechanism which can stretch and retract relative to a fixing frame is installed in the fixing frame. Sensors are evenly installed in the plane, in the same vertical axis direction as an outer fixing ring, of the inner wall of an inner fixing ring. A lug base detection mechanism used for detecting a lug base of the detected steel tie rod or inhaul cable is installed on the fixing frame. The device is used for detecting the application of the steel tie rod or the inhaul cable through a metal magnetic memory detecting method. The inner diameter of a detection cavity can be adjusted immediately according to the diameter changes of the detected steel tie rod or the inhaul cable, and the adaptability is good. A detection gap produced by reducing when the outer fixing ring is expanded can be filled in time, and therefore a reducing detection ring is a complete detection ring both before and after reducing; due to the fact that the sensors are evenly distributed, comprehensive detection can be achieved, and the detection accuracy is greatly improved. The rod section of the detected steel tie rod or inhaul cable is detected, and meanwhile the lug base of the detected steel tie rod or inhaul cable is comprehensively detected.

The invention relates to an intelligent inspection robot which comprises a body. A charging connector is arranged on the front side of the body. Infrared distance measuring sensors are arranged on thefour faces of the body respectively. A host box is arranged on the upper side of the body. Heat dissipation holes are formed in the front side and the rear side of the mainframe box respectively. A first electric telescopic rod is arranged in the middle of the mainframe box. A spherical rotary joint is arranged on the upper side of the first electric telescopic rod. The right side of the spherical rotating wire joint is connected with an infrared thermal imager. The left side of the spherical rotary joint is connected with a camera. An illuminating lamp is arranged on the left side of the camera. A temperature and humidity sensor is arranged on the upper side of the right portion of a host. A sound sensor is arranged on the upper side of the temperature and humidity sensor. Crawler beltsare arranged on the two sides of the lower side of the body. Dusting brushes are arranged on the lower side of the body. A dust suction hole is formed in the rear side of the dusting brushes. A dust collection box is arranged on the rear side of the dust suction hole. The system has the advantages of being long in endurance time, high in intelligence, comprehensive in inspection, diversified in function and comprehensive in detection.

The invention discloses a recording notification fault detection method and device, equipment and a medium. The method comprises the following steps: establishing a virtual dial testing scene under the VoLTE network, configuring calling terminal information and called terminal information required by the virtual dial test scene, and determining the identification information of the standard recording notification corresponding to the virtual dial test scene, wherein one end of the calling terminal and one end of the called terminal are VoLTE terminals, and the other end of the calling terminaland the other end of the called terminal are VoLTE terminals, 2G/3G terminals or IMS fixed network terminals; sending the calling end information and the called end information to a core network; receiving identification information of an actual recording notification returned by the core network according to the calling end information and the called end information; and determining a recordingnotification fault according to the identification information of the standard recording notification and the identification information of the actual recording notification. According to the recording notification fault detection method and device, the equipment and the medium provided by the embodiment of the invention, the recording notification fault under the VoLTE network can be intelligently and comprehensively detected with high efficiency and low cost.

The embodiment of the invention provides a reverse pop detection method and a device thereof. The method comprises steps of determining an abrupt change time domain signal existing in a to-be-detected target audio file and an audio position where the abrupt change time domain signal exists; searching alternative audio files which are the same as the target audio file in label information and durance information; determining a homologous audio file according to a frequency spectrum energy digital sequence of the target audio file and the frequency spectrum energy digital sequence of each alternative audio file; determining a first homologous audio file with the abrupt change time domain signal in the audio position from the homologous files; and determining an abrupt change time domain signal existing in the target audio file to be a reverse pop according to the quantity of first homologous audio files and the quantity of second homologous audio files except for the first homologous audio files. According to the embodiment of the invention, reverse pops can be intelligently, quickly and precisely detected.

The invention discloses a smart city water area garbage density detection method based on a CIM and a block chain. The method comprises the following steps: constructing a city information model of athree-dimensional digital space; establishing a water area garbage density detection deep neural network; training a water area garbage density detection neural network; splitting the water area garbage density detection deep neural network reasoning task into a plurality of sub-reasoning tasks according to the number of network layers, and allocating the sub-reasoning tasks to cloud host nodes; connecting blocks distributed in different cloud host nodes to generate a block chain private chain; performing deep neural network reasoning, and outputting a water area garbage density map and a water area garbage density level; and visualizing the city information model through a WebGIS technology. By means of the method, intelligent smart city water area garbage density detection is achieved, convenience is provided for water area garbage salvage, the detection precision and efficiency are high, the needed cost is low, and data privacy protection in the data processing process is well guaranteed.

The invention provides a plastic product quality detection method, which comprises the following steps: carrying out integrity scanning on a first plastic product through an omnibearing infrared scanner to obtain a first 3D infrared scanning model, the first 3D infrared scanning model being obtained by passing the first plastic product through the omnibearing infrared scanner according to a first posture; obtaining drawing parameter information of the first plastic product; performing 3D modeling according to the drawing parameter information to obtain a standard 3D model of the first plastic product; performing feature comparison on the first 3D infrared scanning model and a standard 3D model to obtain a first similarity value and a first distinguishing feature; obtaining position information of the first distinguishing feature and attribute information of the first distinguishing feature; and inputting the first similarity, the position information and the attribute information into a plastic product quality detection model to obtain a first quality detection result.

The invention provides a device for detecting multiple steel wire ropes and application thereof. The device comprises a bracket main body, wherein a plurality of detection bases are installed inside the bracket main body, each steel wire rope to be detected correspondingly penetrates through the middle of one detection base, a pair of lifting mechanisms is installed on the bracket main body, and can ascend and descend along the steel wire rope to be detected, so that the movement detection of the whole device on the multiple steel wire ropes can be realized. The device is used for detecting the steel wire ropes through a metal magnetic memory detection method. The adjustable device can guarantee the same lifting heights, the multiple steel wire ropes can be detected at a time simultaneously, the reference performance between the detection results is strong, the abrasion of a sensor can be avoided, and meanwhile, the detection precision can be guaranteed. A three-dimensional magnetic memory detection sensor is installed on the bracket, and the multiple steel wire ropes are detected according to the metal magnetic memory detection principle, so that the detection of stress concentration in the early stage and the prediction of the service lives of the steel wire ropes can be realized.

A method for detecting crack width of concrete surface includes such step as spraying a predetermined amount of liquid on concrete surface containing crack, recording liquid absorption process of concrete surface by video frequency in a predetermined time, and intercepting frame photograph at a predetermined time interval in video frequency; Storing a group of frame photographs obtained on the concrete surface into a training set in correspondence with the crack width of the concrete surface; N concrete surface regions containing cracks are sampled according to different absorption rates of liquid by the concrete at different crack widths, and the step A is repeated; B, train that training set with the method of depth learn to obtain the trained model; c, for video record of water spray onthat concrete surface containing cracks to be detected, inputting a group of frame photograph obtained on the concrete surface into the model for detection, and obtaining the crack width of the concrete surface. The invention can quickly, accurately and directly detect the crack width.

The invention discloses an image-based non-contact palm herpes detection system and a detection method thereof. The palm herpes detection system comprises a high-performance processor, an image acquisition device, a communication interface and an upper computer. The image acquisition device obtains palm color image data, the image is transmitted to the high-performance processor through the communication interface. The high-performance processor processes the acquired image, detects skin color contour, screens skin color, judges palm features, detects and recognizes palm herpes spots, and transmits the detection and recognition results to the upper computer. The image-based non-contact palmar herpes detection system and the detection method thereof are convenient to detect, do not rely onartificial medical technology, can quickly detect suspected palmar herpes spots, and effectively prevent the spread of herpes.

The invention relates to an orchard modeling analysis system and method based on unmanned aerial vehicle positioning shooting. The system comprises an unmanned aerial vehicle, a fruit marking module and a virtual modeling module; the fruit marking module is used for obtaining the position of the unmanned aerial vehicle from a positioning module, determining the position of a fruit image on an image shot by a camera module, and marking fruits on the image according to the position of the unmanned aerial vehicle and the position of the fruit image, wherein The virtual modeling module is used forcalculating an unmanned aerial vehicle position range and a fruit image position range in which a target fruit can be shot; and according to the calculated unmanned aerial vehicle position range, thefruit image position range and the unmanned aerial vehicle position and fruit image position marked by the target fruit, calling a plurality of images of the target fruit shot by the camera module, and establishing a three-dimensional model of the target fruit and/or the plant where the target fruit is located according to the plurality of images of the target fruit. According to the method, thefruits in the orchard can be conveniently marked, and the health conditions of the fruits and plants where the fruits are located can be modeled and analyzed.

The invention provides a device and method for detecting the instability of a steel-structure pressing rod. The device comprises a magnetic shielding outer bracket ring, an inner bracket ring, a retractable magnetic memory detection pen and a power walking mechanism; the inner bracket ring is mounted in the magnetic shielding outer bracket ring; the power walking mechanism is mounted on an outer wall of a rear retaining ring at the rear end of the magnetic shielding outer bracket ring; the retractable magnetic memory detection pen is mounted in the inner bracket ring; the steel-structure pressing rod penetrates through the inner bracket ring; the power walking mechanism can walk on the steel-structure pressing rod; the retractable magnetic memory detection pen is used for detecting the steel-structure pressing rod. The retractable magnetic memory detection pen provided by the invention is provided with two three-dimensional magnetic memory sensors and a distance sensor in total and can realize complete and accurate acquisition of normal and tangential magnetic signals through annular free movement of the device; the detection of rod pieces with various types of sections can be realized, and accurate judgment on the stability of various types of members to be detected is guaranteed.

The invention discloses a coal ash content on-line detection system based on deep learning and a detection method thereof, the system comprises a classification model and a regression model, the models are deployed to an embedded device after being trained, and distributed deployment of a coal ash content on-line detection function is realized; and real-time detection of the ash content is completed by a matched hardware image acquisition system. The on-line detection method comprises the following steps: acquiring a coal microscopic image and corresponding ash content through an image acquisition device, establishing a coal microscopic image database, constructing a feature extraction network based on a deep learning method to perform automatic feature extraction on the coal ash content image, designing a classification model and a regression model to complete a final decision, and obtaining an accurate coal ash content prediction result. Compared with other coal ash content detection methods, the method has the advantages of high detection precision, high speed and the like.