6008results about How to "The detection process is fast" patented technology

The invention belongs to the technical field of the pattern identification and image processing, and relates to a license plate detection and identification method based on the maximum stable extremal region, which comprises the following steps: extracting the maximum stable extremal region (MSER) to obtain text regions of candidate license plates; adopting one effective feature for description of each extremal region, classifying the extremal regions into 'text' regions and 'non-text' regions with a classifier obtained in prior training, and extracting a license plate from original images based on the characteristics of the structure of the license plate; and describing with the characteristics of the shape context, and completing the character identification through template matching. Since the maximum stable extremal region has the affine invariance, high stability and multi-scale features, and the regions are determined only according to the gray value and are not light-sensitive,the license plate detection and identification method with the maximum stable extremal region as the base is applicable to the complex background, and has the advantages of good stability and high identification rate.

The invention provides a lane line detection method based on a full-convolution network and relates to the field of traffic information detection. According to the method, probability operation is performed on an output feature map of the full-convolution lane line detection network, the probability that a lane line appears in each region in an input picture is obtained, a prediction probability threshold is set, and extraction and detection of lane lines are realized. Through the method, detection of straight lane lines and curved lane lines can be realized simultaneously; a lane line detection loss function is used to train the full-convolution lane line detection network, so that the detection effect of the lane lines is improved; the convolutional neural network learns abstract features of the lane lines from a lane line classification dataset instead of simply extracting external features of the lane lines; detection of a newly input image can be realized just by storing a lane line detection network model, so that storage space is saved, and the method is suitable for vehicle-mounted embedded equipment; and the small and shallow full-convolution lane line detection network isadopted to perform detection acceleration, so that detection speed is high.

A real-time visual detection and identification method for high speed rail surface defects comprises the following steps of: (1) image acquisition; (2) image preprocessing; (3) defect preliminary detection, that is, performing logic or operation combination of detection results based on gray scale compensation with detection results based on top-hat operation, and detecting whether an abnormal area exists in an image, if not, finishing the detection, or else continuing the processing; (4) defect accurate positioning, that is, accurately positioning the defect by an algorithm of bonding single defect, an algorithm of filling holes in the defect area, and an algorithm of selecting the main defect, and extracting the defect area through marks; (5) defect classification, that is, extracting and selecting characteristics of the defect area, designing and training a BP neural network, and classifying the defects by the BP neural network. The invention has the advantages of simple principle, high automation degree, high detection speed, and high detection precision.

The invention relates to a nondestructive detection method for tea quality based on a near infrared spectrum and a machine vision technology and a device thereof. The detection method comprises the following steps: spectrum and image acquisition to tea are synchronously performed, obtained initial data are subjected to preprocessing, feature extraction and fusion, and then an expert evaluation result is combined to establish a quality grade evaluation model of the tea to quickly and nondestructively detect the comprehensive internal and external quality of the tea. The detection device comprise a spectrum acquisition device, an image acquisition device and a computer, wherein the spectrum acquisition device and the image acquisition device finish the signal acquisition of a tea sample and then transfer a signal to the computer; and the computer finishes the extraction and fusion of feature information, and then substitutes in the pre-established model to judge a final result of the quality of the tea. The method applies a fused nondestructive detection means based on the two technologies to the detection process of the comprehensive quality of the tea, has the advantages of high detection accuracy, good result consistency, strong automation degree and so on, and creates conditions for standardized classification of the quality of the tea.

The embodiment of the invention discloses a training and detecting method of a malicious code family. The method includes the steps that malicious codes are mapped into images, and image features of the images are extracted; the malicious code images are clustered according to the image features, and malicious code family labeling is conducted on different types of malicious code images obtained after clustering; a convolutional neural network model is established; the convolutional neural network model is trained with malicious code image sets obtained after malicious code family labeling, and the trained convolutional neural network model is used as a detection model; family detection is conducted on a malicious code sample to be detected and the variation of the malicious code sample with the detection model. The embodiment of the invention further discloses a training and detecting device of a malicious code family.

The present invention provides a neural network-based face detection model training method, a neural network-based face detection method, a neural network-based face detection model training system and a neural network-based face detection system. The training method includes the following steps that: the loss function of the bias network layer of a prediction face frame is calculated according to the bias information of the prediction face frame relative to a default face frame and the bias information of a real face frame relative to the default face frame; the loss function of the confidence network layer of the prediction face frame is calculated according to the confidence of the default face frame; the error of the two loss functions is calculated, and is fed back to a neural network, so that the weight of the neural network is adjusted; iterative training is repeated until convergence appears, so that a face detection model can be obtained, and therefore, the prediction face frame can contain a face more accurately. The detection method includes the following steps that: a face image to be detected is inputted to a trained face detection model, bias information and confidence are outputted; corresponding prediction face frames are calculated according to the bias information; and a prediction face frame corresponding to confidence greater than a preset confidence threshold or the highest confidence is selected as a face detection result.

The invention provides a testing machine used for testing the appearance of a printing circuit board, comprising a frame which is provided with a transmission mechanism; the front end of the transmission mechanism is provided with an upper plate mechanism; the transmission mechanism is internally provided with two scanning mechanisms; a panel turnover mechanism is arranged between the two scanning mechanisms; the tail part of the transmission mechanism is butt-jointed with a sorting mechanism; and all mechanisms are controlled by control host machines cooperatively. The testing machine adopts a whole-line digital camera photographing technique to detect the defects on the printing circuit such as foreign matter, exposed copper, supplemented oil, scraping, poor gold-plating, character error, non-uniform green oil, non-uniform welding disc, non-uniform depth of channel, residual copper, omitted printing, developing and the like, automatically feeds and sorts the nondefective and waster, improves the simpleness and reliability of the operation, can observe the defects of the board in real-time by adopting a powerful image processing and analysis software system to provide a plurality of image processing and analysis systems, has quick detection speed, high detection efficiency, safe and reliable equipment, simple operation, is convenient and easy to be learned and meets the man-machine interaction.

The invention discloses a method for detecting vehicle door outer plate shape and size based on single-projection encoding structured light and aims to break through the condition that structured light visual detecting technology is rarely applied to online detection on the vehicle door outer plate shape and size. The method comprises the following steps of 1, calibrating a structured light detecting system, i.e. calibrating a camera (2) and a projector (4), and establishing corresponding non-linear relation between a space position and an image coordinate; 2, designing an encoding projection template; 3, decoding and identifying a detecting image, i.e. projecting the encoding projection template to the surface of a vehicle door outer plate by using the projector (4), shooting a surface image of the vehicle door outer plate by using the camera (2), and transmitting the shot image to a computer to be processed and identified; and 4, calculating a space point cloud coordinate and evaluating an error, i.e. 1) calculating a three-dimensional space coordinate of surface projection image characteristic points of the vehicle door outer plate, and 2) matching and rectifying a detecting point cloud obtained through calculation and a point cloud model measured by a three-coordinate measuring machine.

The invention discloses a multi-vision defect detecting equipment and a multi-vision defect detecting method for a large-size LCD glass substrate in a production line. For the equipment, a linear scanning imaging detection system is adopted for acquiring the high-definition gray image of an LCD glass substrate to be detected, and the equipment is simple in structure and convenient to operate; according to the method disclosed by the invention, the acquired image is processed, a kmeans clustering method is adopted to carry out defect existence judgment on the processed image of the LCD glass substrate, and by making the defect regions, the defect category is judged by a classification method of a support vector machine (SVM), and the defect quantity is counted. Processed image and detection results are transmitted to an office in real time by a filed bus control system for operators on duty to check, and the field production parameters are transmitted, so that remote monitoring of the system is realized.

The invention proposes a dangerous driving behavior real-time detection method based on deep learning, and the method comprises the steps: building a front car image collection system, collecting a training sample, carrying out the manual marking, and building a dangerous driving behavior data set; proposing a spatial pyramid pooling convolution depth confidence normalized classification network (SPP-CDBRNet) model according to the characteristics of the dangerous driving behavior data set based on a deep learning method; carrying out the preprocessing of the built data set, carrying out the training of the SPP-CDBRNet model through a random gradient descent method with a momentum and the data set after preprocessing, and obtaining an SPP-CDBRNet which precisely recognize whether there is a dangerous driving behavior (behaviors of using a cellphone and smoking during driving); carrying out the detection of a front car image through the determined SPP-CDBRNet model, and achieving the real-time detection of the dangerous driving behavior. The method can effectively improve the detection precision of dangerous driving behaviors, is good in instantaneity and mobility, and is good in application prospect.

The invention discloses visual detection equipment of a circuit board. The visual detection equipment of the circuit board comprises a bearing plate and a control system, a conveyor trough arranged on the bearing plate, a conveyor belt clamped in the conveyor trough and used for conveying a supplied to-be-detected circuit board, a first visual detection device, a second visual detection device, a finished product collection device and a defective product collection device, wherein the first visual detection device comprises a first CCD (charge coupled device) imaging system and a first blocking mechanism; the second visual detection device comprises a second CCD imaging system and a second blocking mechanism; the finished product collection device is used for collecting finished circuit boards after detection; and the defective product collection device comprises a defective product rejecting manipulator and a defective product collection conveyor belt used for collecting defective products. Compared with the prior art, only one conveyor belt is adopted for the detection process, so that the cost is saved greatly, and the visual detection equipment of the circuit board can perform detection on a first surface and a second surface of the circuit board respectively, the detection covering range is large, and the detection speed is high.

The invention relates to a multi-channel portable colloidal gold test paper strip quick test device, which mainly comprises a detector, wherein the detector comprises a detecting part, an electrical part and a software detecting part; the detecting part comprises a plurality of mutually independent detecting channels; the electrical part comprises an ARM data processor, a keyboard and an LCD; image data detected by the detecting part are processed and displayed by the electrical part; and the software detecting part controls detecting steps. The device has a multi-channel detecting structure, and four channels can carry out the same detection item as well as different detection items. The device has the advantages of high detection speed, large amount of samples and great practical value.

The invention relates to an on-line detection grading device for a lithium ion battery pole piece and a method thereof. In the device, image data of the lithium ion battery pole piece are acquired by using an embedded visual recognition device and are analyzed and processed to judge whether the appearance quality of the pole piece meets requirements or not, so that the actions of a turnover part and a grading part are controlled to realize real-time on-line double-side detection and grading of the electrode pole piece. The method has high detection speed; generally, the detection of the frontside and the back side of one electrode pole piece can be finished within 1 second, so that the production efficiency is increased; if the pole piece is manually detected, longer time is required to be taken; the detection accuracy is high; the embedded visual detection technology is adopted, so that the accuracy is high and the stability is also realized, and a detection result has high uniformity; the running time is long, namely a detection system can uninterruptedly run for 24 hours, so that the productivity is improved and the labor cost is saved; and on-line detection can be realized, so that the on-line detection of the electrode pole piece is realized, the interference in a manual operation process is reduced, and the productivity is improved.

The invention provides an SAR image target detection method based on the convolutional neural network and mainly solves the problem of low detection speed and low detection accuracy rate in the conventional SAR image target detection technology. The method comprises the steps of firstly training a convolutional neural network classification model M0 and designing two to-be-trained convolutional neural network models M1 and M2 based on the model M0; secondly, using an expanded MiniSAR training set as the input for the to-be-trained convolutional neural network models M1 and M2 and acquiring an optimal slice extracting model M12 and an optimal detection model M22 through training; thirdly, detecting the MiniSAR training set by using the optimal slice extracting model M12 and the optimal detection model M22; fourthly, reserving candidate areas with output probability values of the optimal detection model M22 being greater than a threshold value and acquiring a detection result. The method has the advantages of high detection speed and high detection accuracy rate and can be used for vehicle target detection.

The invention relates to a microfluidic control chip which is formed by bonding and sealing PDMS (Polydimethylsiloxane) material integrated with a micro valve, and comprises an upper layer of air control passage, a middle layer of PDMS membrane and a lower layer of microfluidic passage. A real-time PCR virus quick detection device comprises the microfluidic control chip, a temperature control unit, a signal detection unit and a data processing and control unit. The invention designs the microfluidic control chip which is integrated with the micro valve, is in a three-layer PDMS structure, andis used for virus detection and the real-time fluorescent PCR detection device. Compared with the traditional method, the invention has the advantages of short time, less sample usage, quick detection speed, simpleness and convenience in operation, integration and the like.

The invention relates to a direct power-spectral method-based all-weather long drainage basin water quality monitoring and early-warning system, belonging to the technical field of water quality monitoring. The system comprises a plurality of water quality monitoring devices, a cloud data processing platform and a remote monitoring platform; the water quality monitoring devices are placed at a shoreside of a water body, a water surface buoy or other water surface platforms according to needs, used for extracting water quality data of the long drainage basin, use a direct power-spectral method to acquire an absorption spectrum of a monitored water body and sends the monitored data and water quality information to the cloud data processing platform in a wireless transmission manner; the cloud data processing platform summarizes and analyzes the received data and stores into a database; the remote monitoring platform inquires and displays the data stored in the database and realizes the real-time monitoring and early-warning for an all-weather long drainage basin water quality environment. The system can realize the online measuring, real-time monitoring, emergent early-warning, drainage basin analyzing and water environment monitoring of all weathers of the water body in the drainage basin, thus reducing the loss caused by pollution to the water body and ensuring drinking water safety.

The invention belongs to the technical fields of biochips and diagnostic reagents, and discloses a method for detecting various respiratory viruses, and primers and probes thereof. In the invention, nucleotide sequences of 14 respiratory viruses, namely adenovirus, human metapneumovirus, influenza virus A, influenza virus B, respiratory syncytial virus, bocavirus, rhinovirus, coronavirus (HKU1, NL63 and SARS), and parainfluenza virus (type I, type II, type III and type IV) are analyzed, and corresponding reverse transcription primers, PCR primers and specific probes are designed. Specific gene segments are amplified by reverse transcription and multiple asymmetric PCR methods; a fluorescence-coded microsphere group coupled with the virus specific probes and the PCR amplification product are incubated and hybridized by liquid phase chip technology; and finally the Bio-PlexTM200 is used for detection. The detection method has the advantages of high flux, high specificity and sensitivity, stable results and good repeatability, the detection method is easy to operate, and the detection speed is high.

The invention relates to impulse response analytical test apparatus and method for detecting deformation of transformer winding, belonging to the technical field of transformer winding detection. The apparatus mainly comprises an impulse generator, a broadband digital oscilloscope, a computer, a high-voltage probe, a head end signal impedance matching device, a tail end signal impedance matching device and a screened coaxial cable. In the method, site test and analysis are carried out by using the apparatus. The invention has the advantages of high test sensitivity, rapid detection, strong capacity of resisting disturbance and flexible analytical method, and can detect slight deformation of the transformer winding and judge early potential faults effectively with high judging accuracy. The invention can be widely applied to the detection of the deformation of the transformer winding and the state overhauling.