69results about How to "Fast detection rate" patented technology

The embodiment of the invention discloses a method for detecting a human face. The method comprises the following steps of: carrying out LBP (Length Between Perpendiculars) characteristic calculation on an acquired image to obtain an LBP characteristic pattern; distinguishing the LBP characteristic pattern by using an AdaBoost classifier trained based on LBP characteristics to obtain first human face region data; carrying out Haar characteristic solving on the first human face region data to obtain an Haar characteristic pattern and distinguishing the Haar characteristic pattern by using the AdaBoost classifier trained based on the Haar characteristics to obtain second human face region data; clustering the second human face region data and combining the overlapped human face regions to obtain fourth human face region data; and outputting a human face region in the image. The embodiment of the invention also discloses a device for detecting the human face. According to the method and the device disclosed by the invention, the detection rate of human face detection is greatly increased; the false detecting rate is reduced, high detection speed is obtained, an interframe human face tracking step can also be added and further the detection speed is increased and the requirements of users on high detection rate and high detection speed are met.

The invention provides a method for rapidly and nondestructively detecting astaxanthin in C.zofingiensis cells. The method comprises the following steps of (1) re-suspending algal cells collected from a known C.zofingiensis suspension sample to regulate the known C.zofingiensis suspension sample into heavy suspension with cell density of 6*10(5) to 2*10<6>CFU / mL by virtue of de-ionized water or a phosphate buffer solution, detecting the average fluorescence intensity of the heavy suspension in an FL1 channel or an FL2 channel by virtue of a flow cytometer, and performing linear regression on a log value of the content of astaxanthin in the known C.zofingiensis suspension sample and a log value of the corresponding measured average fluorescence intensity in the FL1 or FL2 channel to obtain a standard curve; (2) preparing a sample to be detected: re-suspending algal cells collected from C.zofingiensis suspension to be detected to regulate the C.zofingiensis suspension to be detected into algal cell heavy suspension to be detected with cell density of 6*10(5) to 2*10<6>CFU / mL by virtue of de-ionized water or a phosphate buffer solution; (3) detecting the average fluorescence intensity of the sample to be detected in the FL1 channel or the FL2 channel by virtue of the flow cytometer, and calculating the content of astaxanthin in the sample to be detected according to the standard curve obtained in step (1). The method has the characteristic of high detection efficiency.

The invention discloses an LED light source short-circuit detection method and device, an LED backlight and a liquid crystal display device. The LED light source short-circuit detection method and device aims to simplify the short-circuit detection of an existing LED light source. The LED light source short-circuit detection method is applied to the backlight of the liquid crystal display device, wherein the backlight comprises a plurality of LED lamp strips. The LED light source short-circuit detection method comprises the following steps of inputting M unequal detection voltages to each LED lamp strip; receiving and recording breakover feedback of each LED lamp strip at under each detection voltage; judging whether LED lamps are short-circuited in each LED lamp strip according to the breakover feedback, wherein M is an integer which is not smaller than 1, and the maximum detection voltage is not larger than the normal working voltage of each LED lamp strip. The LED light source short-circuit detection method and device is used for the short-circuit detection of the LED light source, and has the advantages of being easy and convenient to achieve, fast in achievement, sensitive in detection and the like.

The invention relates to an augmented reality image data processing method and device, computer equipment and a storage medium. The method comprises the steps of obtaining a reality scene image collected by a camera carried on a refrigerator; identifying food materials in the real scene image through a food material identification model, wherein the food material identification model is a cascadenetwork of a feature pyramid network and a YOLOV3 network; obtaining virtual resources corresponding to the food materials in the real scene image, wherein the virtual resources are virtual animationresources designed according to nutritional ingredients of the food materials; and displaying the virtual animation resources and the real scene images in a combined manner, wherein the feature pyramid network and the YOLOV3 network can extract food materials of different scales; according to the invention, the requirements of various food materials are met, the detection speed is high, the food materials can be quickly and accurately recognized, the virtual animation image is displayed in an AR (augmented reality) mode to display the nutritional ingredients of the food materials, a user selects the food materials according to the nutritional ingredients, and the diet quality is guaranteed.

Provided is a method for detecting the saturation of a current transformer (CT) on the basis of a third-order central moment. The method comprises: sampling current at the secondary side of the CT; acquiring a current waveform from a position near the peak value of a waveform by using a data window in order to obtain a waveform symmetric about the peak value and with less sampling points; computing a dispersion coefficient by using the third-order central moment after the acquired waveform is subjected to waveform mirror image transformation; and outputting a detected result. In the method, a dispersion coefficient after a standard sinusoidal current waveform is subjected to mirror image transformation. By means of analysis and computation, the dispersion coefficient can be considered as a fixed value so as to be used as a comparison basis. The method mainly solves problems of low detection rate and low detection precision when the CT saturation occurs, may fast and accurately detect the generation of the CT saturation just by using sampled data in half cycle so as to improve relay protection reliability to a certain extent and have certain practical significance in the aspect of guaranteeing the accurate and reliable operation of the CT.

The embodiment of the invention provides a method for detecting proxy network access. The method comprises the following steps: establishing a proxy device identifier table in advance; receiving a Web webpage request message sent by a first user terminal after establishing a TCP connection with a Web server; feeding back a message used for detecting identification information of the user terminal to the first user terminal according to terminal type information in the Web webpage request message; receiving a message containing the identification information of the first user terminal sent by the first user terminal; when the source IP address of the message containing the identification information of the first user terminal is in the proxy device identifier table, determining whether the user terminal is a user terminal that surf the Internet through a proxy device according to the identification information of the first user terminal and the proxy device identifier table. Meanwhile, the invention further provides an apparatus for detecting proxy network access. According to the method and apparatus provided by the embodiment of the invention, the false alarm rate is low, and the network access experience of users is good.

The invention discloses an infrared image small target detection method based on a generative adversarial network. According to the method, a false background image block is reconstructed for a targetimage block containing a small target; pairing the false background image block with the target image block to obtain first pairing data, training the generative adversarial network by using the data; pairing the two original background image blocks which do not contain the small target in the image sequence block to obtain second pairing data; continuously training by using the data to obtain atarget generative adversarial network model; and inputting a to-be-detected image block sequence into the target generative adversarial network model to obtain a detection sequence result, splicing the detection sequence result according to the size of the to-be-detected original infrared image, making a residual error between the spliced image and the to-be-detected original infrared image, and filtering the image after the residual error is made to obtain a detection result. Through the method provided by the invention, background interference can be greatly suppressed while a relatively high detection rate is ensured.

The invention discloses a magnetic nano composite material and a preparation method and application thereof. The magnetic nano composite material is of a core-shell structure, an inner core is Fe3O4, a shell layer is SiO2, and a middle layer is further arranged between the inner core and the shell layer; wherein the intermediate layer is from a polymer; and the polymer is selected from at least one of polyvinylpyrrolidone and polydopamine. According to the magnetic nano composite material provided by the invention, sample DNA can be rapidly extracted, and the extracted DNA is applied to a qPCR (real-time fluorescence quantification) detection technology.

The invention belongs to the technical field of deep learning anomaly detection, and discloses an industrial part surface defect detection method and system based on an anomaly detection algorithm. The method comprises the following steps: obtaining an image of a to-be-detected product, and collecting the surface image data of a to-be-detected good product shot by an industrial camera; the image is input into a teach-student anomaly detection network model to be trained; detecting defect data and non-defective product data by using a model produced after training is completed; returning a detection result to the client, and displaying the detection result by the client; and transmitting the detected abnormal region and the size of the region to a client of software, and judging whether the defect needs to be eliminated or not through an area threshold value on the software. The detection of surface abnormal areas of most parts with simple surface structures can be met. Compared with target detection, a large amount of model training and data collection time is saved, and the method has a great promotion effect on project promotion and acceptance as soon as possible.