69results about How to "Accurate detection and identification" patented technology

The invention discloses a video-based detection and recognition system and a method of vehicles. The system comprises a target characteristic database module, a moving target detection module and a vehicle recognition module. A vehicle sample target characteristic database is built to store data by virtue of the target characteristic database module. After a moving target zone is detected by the moving target detection module, the vehicle recognition module builds a search window and recognizes vehicles. The system and the method resolve the technical problems of poor vehicle distinguishing capability and low distinguishing accuracy of the video-based vehicle detection under the condition of obvious background movement, vehicle adhesion, vehicle occlusion and the like in the prior art. The video-based detection and recognition system and the method of the vehicles have the advantages of excellent vehicle distinguishing capability and high accuracy.

The invention discloses an orchard bactrocera dorsalis insect damage recognition system based on digital signal processing (DSP) and the internet of things. The orchard bactrocera dorsalis insect damage recognition system comprises an installation support, a novel insect trap fixed on the installation support, an industrial camera arranged above the novel insect trap and a casing fixed on the installation support. A photoelectric detection module is arranged in the novel insect trap; a DSP digital image processing module, a scanning tunneling microscope (STM) 32 functional circuit, a global system for mobile communications (GSM)/general packet radio service (GPRS) wireless communication module and a charging control module which are sequentially connected are arranged in the casing; a displayer and a touch screen are arranged on the surface of the casing; and the charging control module is further connected with a solar panel. According to the orchard bactrocera dorsalis insect damage recognition system, the DSP technology, the machine vision technology and the internet of things technology are combined, the machine vision technology is achieved on a DSP processor platform and applied to detection and recognition of orchard bactrocera dorsalis, and accordingly fast and accurate detection and recognition of orchard bactrocera dorsalis are achieved. Simultaneously, automatic real-time monitoring of orchard insect damage is achieved and cost of manpower and material resources in orchard insect damage monitoring is reduced.

The invention relates to method for a synchronous detection of sweet potato feathery mottle virus (SPFMV), sweet potato virus C (SPVC), sweet potato virus G (SPVG) and sweet potato virus 2 (SPV2), and belongs to the field of plant protection. The method comprises the following steps of respectively designing and synthesizing specific forward primers and a universal reverse primer of SPFMV, SPVC, SPVG and SPV2, extracting total RNA of tissue of a diseased plant through a cetyltrimethyl ammonium bromide (CTAB) method, and carrying out one step quadruplex reverse transcription-polymerase chain reaction (RT-PCR) magnification process to realize a synchronous detection of SPFMV, SPVC, SPVG and SPV2. Primers designed by the method have strong singularities and SPFMV, SPVC, SPVG and SPV2 share areverse primer thus an interaction of primers is reduced. In the invention, a CTAB method is utilized for extracting RNA from tissue of an RNA virus infected plant, thus a quality of RNA is guaranteed and a detection cost is reduced effectively; and the inverse transcription and the multiple PCR are completed in one step thus a detection time is saved. The method has the advantages of rapid detection speed, high efficiency, strong singularity, high sensitivity and low cost and can realize a synchronous detection of four kinds of sweet potato virus thus has wide application prospects.

The invention discloses a power terminal loophole attack detection method based on message features, and belongs to the technical field of intelligent power grid terminal equipment safety. The methodcomprises the following steps: S01, acquiring communication message data between power terminal equipment and a master station, and classifying the communication message data of the power terminal equipment in a normal working state and an attacked state into a positive sample and a negative sample; s02, performing feature extraction on the positive sample and the negative sample, and forming a sample feature vector; s03, based on the sample feature vector, selecting a classifier to perform deep neural network training, and generating a vulnerability attack detection model; s04, collecting real-time communication message data between the power terminal equipment and the main station in work; s05, performing feature extraction on the real-time communication message data, and forming a detection feature vector; and S06, inputting the detection feature vector into a vulnerability attack detection model to detect whether the power terminal is attacked or not and the attack type. Accordingto the method, the power terminal equipment is subjected to safety monitoring from a network layer, and the power grid safety is improved.

The invention provides a human body target identification method and apparatus. The method comprises the steps of obtaining a depth image; extracting image pixel characteristics in the depth image; inputting the image pixel characteristics into a human body deep learning model to perform identification and classification; judging whether categories of the image pixel characteristics are matched with existing human body part tags in the human body deep learning model or not; and if the categories of the image pixel characteristics are matched with the existing tags in the human body deep learning model, outputting the tags corresponding to the pixel characteristics. According to the method and the apparatus, the image pixel characteristics are identified by adopting the deep learning model, and human body target detection and identification are finished, so that the detection and identification process is simplified and the detection and identification efficiency is improved.

The invention discloses a method for testing and identifying cepaea hortensis through polymerase chain reaction (PCR), and the method comprises the following steps that a specific primer is applied to identify the molecular features of cepaea hortensis; a forward primer of the PCR is CN-(P1):5'-ACCTCCTTCCTTTCTACT-3', and a reverse primer is CN-(P2):5'-GTCAACATCTATCCCAAC-3'. The total volume of a PCR system is 25mu.L, including 12.5mu.L of 2*TaqPCRMasterMix, respectively 0.5mu.L of the forward and reverse primers, 2mu.L of a deoxyribonucleic acid (DNA) template, and the rest is sterilizing ddH2O; and PCR procedures are as follows: predegeneration is carried out for 5 minutes at 5DEG C, 50 seconds at 95DEG C, 30 seconds at 52DEG C and 50 seconds at 72DEG C, and the process is repeated for 35 times; and extension is carried out for 10 minutes at 72DEG C, and the reaction is over. According to the method for testing and identifying cepaea hortensis through PCR, the cepaea hortensis can be quickly and accurately tested and identified, a novel technical means is provided for the testing and the identification of the dangerous and harmful creature cepaea hortensis, and important significance in preventing the cepaea hortensis from being spread and diffused is realized.

The invention relates to an illegal website identification system and method based on a critical path. A system framework is divided into four layers, namely a user layer, an application service layer, a technical support layer and a data storage layer. The user layer provides a main account of the system; the application service layer provides a main function module of the system; the technical support layer comprises a relevant tool used in the system development process and a core algorithm program; the data storage layer provides data used in the system. According to the detailed functional division, the function module of the illegal website identification system based on the critical path comprises data preprocessing, website similarity calculation, website clustering, illegal website critical path extraction and illegal website identification. The system develops a similarity calculation program based on the Path through taking URL self characteristics as research start points, can accurately calculate the similarity among websites and obtain the effective URL critical path based on the website similarity and a Fast Unfolding clustering algorithm, and finally can discover illegal websites in unknown websites through the URL critical path.

The invention provides a method of detecting and identifying cernuella virgata da costa through a fluorescent quantitation PCR. Molecular characteristics of cernuella virgata da costa can be identified by using a specific primer and a probe, an upstream primer of the fluorescent quantitation PCR is CVF:5'-CGTTTCGGGTGGGAGGGTTA-3', a downstream primer of the fluorescent quantitation PCR is CVR:5'-AAGTCGTCCGAGCTGTTCCG-3', the fluorescent probe is CVP:5'-FAM-AACCCAGGGCGCCTCCGCGG-BHQ-3', wherein a fluorescent reporting group which is marked by a 5' end is FAM, and a fluorescent quenching reporting group which is marked by a 3' end is BHQ. The method can detect and identify cernuella virgata da costa quickly and precisely, the method has the advantages of high specificity and sensitivity, short consuming time, and precise and convenient result determination, PCR post processing is not needed, and a false positive and cross pollution can be avoided effectively.

The invention discloses a sparse representation-based underground hidden crack filler identification method. In combination with the feature extraction advantages of empirical mode decomposition (EMD) and wavelet decomposition, firstly, the EMD on the echo signals of different underground crack fillers is conducted to obtain an optimal empirical mode component. Secondly, the original echo signals are subjected to wavelet decomposition to obtain an optimal wavelet decomposition coefficient. Finally, the optimal empirical mode component and the optimal wavelet decomposition coefficient are subjected to correlation analysis, and then the optimal wavelet decomposition coefficient largest in correlation with the optimal empirical mode is adopted as a base of a novel over-complete dictionary. After the over-complete dictionary is obtained, a to-be-identified signal is processed based on the same method described as above. In this way, the optimal feature component of the signal can be obtained. At last, the over-complete dictionary and the to-be-identified optimal component are substituted into the compressed-sensing reconstructing algorithm, so that the signal is reconstructed and identified. The method has the advantages of smaller sampled data volume and higher precision.

The invention discloses a method for detecting and identifying Cernuella virgata Da Costa by PCR. The method comprises identifying molecular characteristics of the Cernuella virgata Da Costa by specific primers, wherein the upstream primer of PCR reaction is HP- (P1): 5'-GTTATTGTTACTGCTCACGC -3', the downstream primer is HP- (P2): 5'-CTGCTAAGACAGGGAGAGAT-3'. The total volume of the PCR reaction system is 25 muL, wherein the volume of 2*TaqPCRMasterMix is 12.5 muL, the volume of the upstream and downstream primers are respectively 0.5 muL, the volume of DNA template is 3 muL and the rest volume is supplied with sterile ddH2O; PCR reaction program comprises carrying out a denaturation at 94 DEG C for 5 minutes at 94 DEG C for 30 seconds, at 46 DEG C for 30 seconds, at 72 DEG C for 1 minute, circulating for 25 times, extending at 72 DEG C for 10 minutes and completing the reaction. The Cernuella virgata Da Costa can quickly and accurately detected and identified by the method disclosed by the invention. The method provides a new technical means for the detection and identification of the Cernuella virgata Da Costa in the 'List of Quarantine Pests of the People's Republic of China of the Entry of Plants' and is of great significance for preventing the Cernuella virgata Da Costa from spreading and diffusing.

The invention discloses a traffic sign recognition method and device. The method comprises the following steps: acquiring image data from a camera; Detecting a region of interest of a current frame ofthe image data by using a deep learning technology to obtain an initial traffic sign detection result; Tracking the traffic sign detection result to obtain a final traffic sign detection result whichcan be a tracking result; And predicting the region of interest of the next frame according to the tracking result.

The invention discloses a method for detecting and identifying Helix aspersa MUller through PCR (polymerase chain reaction). The method comprises the step of identifying the molecular feature of the Helix aspersa MUller by using specific primers, wherein the upstream primer of PCR is HP-(P1): 5'-ACTCTTTGGTATTTGATGCG-3', and the downstream primer is HP-(P2): 5'-ACCTGCTAAAACGGGTAATG-3'. The method can be used for quickly and accurately detecting and identifying the Helix aspersa MUller. The method provides a new technical means for detection and identification of the Helix aspersa MUller in Imported Plant Quarantine Pest Directory of the People's Republic of China, and has significance for preventing the propagation and diffusion of the Helix aspersa MUller.

The embodiment of the invention discloses a target tracking method. The method comprises the steps of obtaining a histogram characteristic vector of a current frame in a first gradient direction; according to the histogram characteristic vector in the first gradient direction, determining the position of a first target window of the current frame, wherein the first target window is a target position determining window in the current frame; acquiring the position of a second target window in a previous frame; according to the position of the first target window and the position of the second target window, tracing the motion track of a target. The embodiment of the invention also discloses a target tracking device.

The invention provides an obstacle identification method and system based on 3D laser point cloud. The method comprises the following steps: acquiring three-dimensional environment laser point cloud data; performing ground segmentation processing on the three-dimensional environment laser point cloud data to obtain non-ground laser point cloud data; partitioning the non-ground laser point cloud data, and marking obstacles in each partition by using two-dimensional Euclidean clustering; and performing packaging processing on the clustering result by using a cuboid model to obtain the position and the size of an obstacle. According to the invention, the real-time performance of ground segmentation and point cloud clustering is effectively improved, and the identification precision of obstacles at a relatively long distance is improved.

A method of and apparatus for on-line detection and identification of synchronous and asynchronous piecewise-constant signals corrupted by noise and distortion. The method achieves joint temporal and amplitude detection and identification by following an initial amplitude approximation with a change detection that continually converges to an accurate joint identification while continually correcting intermediate approximation errors via analysis of temporal signatures. In turn, the temporal identifications employ accumulated statistics partly determined by the initial amplitude approximation.

The invention provides a marine engineer abnormal behavior recognition method based on human body joint point detection and a D-GRU network. The method comprises the steps of designing a human body detection model based on human body joint point detection and a GRU network structure; due to the fact that a cabin is dark in light and complex in working environment, detecting and tracking human body joint points by the model through a joint point detection model, and detecting the abnormal behavior of the marine engineer by the human body joint point detection model; and outputting coordinates of the obtained heat maps of the key points, and inputting the coordinates into a D-GRU network structure to carry out human body posture prediction so as to realize recognition of abnormal behaviors of marine engineers. According to the technical scheme, the problem that in the prior art, when working personnel such as marine engineers work in far places such as ship buildings, engine rooms and boiler rooms, the personal safety of the marine engineers cannot be known and reflected in time is solved.

The invention discloses a binder clip defect detection method and system based on deep learning, and belongs to the technical field of image processing, and the method comprises the following steps: S1, constructing a sample data set; S2, obtaining a pre-training model; S3, constructing a binder clip defect detection and identification model; and S4, carrying out detection and identification. The YOLOv4-Tiny network serves as a basic network, weight parameters in the pre-training network are shared through transfer learning, a global space attention mechanism module is added, the feature representation capacity is improved, and detection and recognition of binder clip defect images can be accurately achieved; and in an image defect detection effect comparison experiment, the method has the advantages that a good detection effect is shown, the mAP value index reaches 91.66%, detection and recognition of the binder clip defect image are achieved, a foundation is laid for follow-up implementation of binder clip detection and a feedback system, and the method is worthy of being popularized and used.

The invention discloses a face detection model training method and device, electronic equipment and a storage medium. A training sample set is obtained; wherein the training sample set comprises at least three types of training samples: a face correct labeling type, a face abnormal labeling type and a face non-labeling type; performing training processing on the first to-be-trained face detection model and the second to-be-trained face detection model based on the training set sample set to obtain a first initial face detection model and a second initial face detection model; wherein the first to-be-trained face detection model and the second to-be-trained face detection model have the same model structure, and index parameters of all neural network layers in the model structures are different; and carrying out joint training on the first initial face detection model and the second initial face detection model based on the retraining sample set to obtain a target face detection model, and carrying out detection processing based on a face image in the target face detection model image. The method achieves the accurate recognition of the image under the condition that the image quality is poor.

The invention discloses a gear defect visual detection method and system based on an improved YOLOv5 network. The system comprises a controller with a built-in improved YOLOv5 network model, a conveying system composed of a first conveying belt and a second conveying belt, and an image acquisition module and a rejection mechanism which are built on the conveying system. According to the improvement mode, an unimproved YOLOv5 network is trained through a sample data set to obtain weight parameters, a convolution attention mechanism module and a repeated weighted bidirectional feature pyramid network are added to a YOLOv5 network model, and the weight parameters are migrated to the improved YOLOv5 network model; and training an improved YOLOv5 network model through the data set, completing the construction of a gear defect detection model, collecting an image through an image collection module, inputting the image into the gear defect detection model for recognition, and rejecting a corresponding defective gear according to a recognition result. According to the invention, accurate identification of gear characteristic defects and automatic detection and sorting of gear multi-surface defects are realized, and the detection efficiency is improved.

Methods, systems, and devices are disclosed for detecting encrypted Internet Protocol packet streams. A probe stream having a known observable parameter is generated. The observable parameter exhibits a known characteristic of a known type of data encrypted within a stream of packets. The probe stream is communicated to a network element via a communications network. When the probe stream is received by the network element, the network element can compare the known observable parameter to an actual value. Any difference between the known observable parameter and the actual value can be used to correct for network-induced variation and other effects, thus ensuring accurate detection and identification of data types within encrypted streams of packets.