47results about How to "Distinctive features" patented technology

The invention relates to the technical field of medical data processing, in particular to an advanced nasopharynx cancer treatment effect prediction system based on deep learning. According to the invention, a data acquisition module acquires multi-modal medical images and clinical data of a patient from a hospital information system; the data preprocessing module is used for cleaning, registering, normalizing and standardizing the data; the feature extraction module extracts image features by using a convolutional neural network and fuses the image features with clinical features to form a comprehensive feature vector; the model construction and training module constructs a deep learning model, and is adjusted and optimized through training; the treatment prediction module inputs the comprehensive feature vector to obtain a treatment effect prediction result; and the result evaluation and feedback module compares prediction and actual results, calculates indexes and collects and feeds back the improved model. The system comprehensively integrates data, accurately mines features, constructs an effective model, provides personalized prediction to assist clinical decision, continuously improves the model performance, and has important significance in improving the prediction accuracy of the treatment effect of advanced nasopharynx cancer, promoting personalized medical treatment and the like.

The invention relates to the technical field of RFID chip manufacturing, in particular to an RFID chip image anomaly detection method and system based on intelligent learning. The method comprises the following steps: acquiring a real-time RFID chip sampling image set, performing feature alignment based on a registration template, generating an average grey-scale map by using an alignment image set, and obtaining a chip edge mask image set by using an edge search algorithm; segmenting the edge mask image set of the RFID chip to obtain an effective edge mask variance image set of the RFID chip; integrating the variance image set to obtain a pre-generated variance image; acquiring field environment data of the RFID chip and performing sensitivity coefficient estimation to obtain an edge sensitivity coefficient; and performing convolution filtering smoothing processing on the pre-generated variance image to obtain a chip edge variance image, and constructing an RFID chip image anomaly detection model based on the edge variance image and the average grey-scale image. According to the invention, the detection accuracy and stability are improved.

The invention relates to a monocular vision tower span measurement method based on RTK correction, and belongs to the technical field of space distance measurement. Comprising the following steps: acquiring ambient light intensity and quality in real time based on an environment where a power transmission and distribution tower is located, and obtaining ambient light adaptation parameters by analyzing ambient light data and adjusting ISO sensitivity and shutter speed of a camera. According to the method, real-time environment light data analysis and camera parameter automatic adjustment are introduced in the monocular vision tower span measurement process, the image quality is effectively improved, tower recognition is more accurate, background interference is reduced, the geographic position and the shooting angle of a shooting point are corrected through the RTK technology, and the measurement accuracy is improved. And perspective transformation correction is carried out on the image in cooperation with a perspective projection theory, and the accuracy of distance measurement data is improved to centimeter level.

The invention discloses a double face recognition area early warning safety protection method, and relates to the technical field of face recognition, a face recognition camera, a database, a background server and a client are included, and the method has the advantages that under strong light, an intelligent light adjustment algorithm can automatically reduce the exposure, improve the contrast and enhance the color saturation; the face features are highlighted, the overexposure phenomenon of the face is avoided, under the backlight condition, the backlight compensation algorithm is started, the overall brightness of the face is improved by increasing the edge brightness and adjusting the hue, the accuracy of feature extraction is ensured, and in addition, based on the face detection technology of the deep neural network, through deep training of massive face image data, the accuracy of feature extraction is improved. According to the invention, the camera can learn the face features in various complex environments, and even if the face is partially shielded and is located at different angles or under poor light conditions, the system can accurately carry out detection and recognition, so that the stability and accuracy of face recognition are greatly improved.

The invention relates to the technical field of image data processing, in particular to a printing scraper damage online detection method. The method comprises the following steps: acquiring a surface image of a scraper to be detected, and for a target pixel point in the surface image, determining a parameter value of the target pixel point according to pixel values of other pixel points in a red channel in a neighborhood range of the target pixel point; adjusting a preset blurring radius according to the parameter value of the target pixel point to obtain a target blurring radius corresponding to the target pixel point; and graying the surface image to obtain a surface grayscale image, performing unsharp mask USM enhancement processing on the surface grayscale image by using the target fuzzy radius corresponding to the target pixel point to obtain an enhanced image, and performing threshold segmentation on the enhanced image to obtain a rust region in the enhanced image. According to the technical scheme, the detection difficulty of damage caused by rusting of the scraper can be reduced.

The invention discloses a permanent magnet generator air gap eccentricity fault identification method and system, relates to the technical field of permanent magnet generators, and uses wavelet transform to extract time-frequency features and generate multi-band feature vectors, so that the system can more sensitively identify subtle features of air gap thickness deviation. And the extraction and analysis precision of characteristic signals of different frequency bands is effectively improved. And secondly, analyzing the air gap thickness deviation condition based on the feature vector, constructing an eccentricity degree index, and quantifying the eccentricity degree, so that the severity of the air gap eccentricity fault of the permanent magnet generator can be intuitively judged. And when the eccentricity degree exceeds a set threshold value, the system automatically sends a fault type checking instruction, so that the fault detection process is more intelligent. Afterwards, abnormal changes of all positions are analyzed through infrared thermal data and temperature gradients, air gap foreign matter faults possibly causing air gap eccentricity are further recognized, and the limitation of single signal analysis in traditional detection is made up.

The invention belongs to the field of artificial intelligence and financial science and technology, and relates to an image small-range modification recognition method, which comprises the steps of forming a to-be-recognized image pair by a to-be-recognized image in a to-be-recognized image library and each reference image in a similar image set, and performing pixel partitioning on the to-be-recognized image pair to obtain a pixel block pair set, calculating the pixel difference of each pixel block pair in the pixel block pair set based on difference discrete cosine transform to obtain a block DCT difference image; performing high-pass filtering based on the block DCT difference image to obtain a high-frequency block DCT difference image; and recognizing a modification area of the to-be-recognized image according to the high-frequency block DCT difference image. The invention further provides an image small-range modification recognition device, computer equipment and a storage medium. In addition, the invention also relates to a block chain technology, and the to-be-identified image library can be stored in a block chain. The identification efficiency and accuracy are improved.

The invention relates to the technical field of railway clearance measurement, and provides a railway clearance automatic measurement method and system based on laser point cloud.The method comprises the steps that in the railway clearance measurement process, overall point cloud data and local point cloud data are collected; marking a first adjacent acquisition moment and the target point cloud data, obtaining an external highlight weight and an adjacent curvature of the target point cloud data, and obtaining a local curvature ratio of the first adjacent acquisition moment; according to an included angle between vectors formed by the corresponding point cloud data in the adjacent acquisition moments and an external highlight original weight of the target point cloud data, obtaining a local height included angle ratio of the first adjacent acquisition moments, and obtaining an optimal orbit length of the first adjacent acquisition moments; and performing railway segmentation processing according to the optimal track length at all acquisition moments in the railway gauge measurement process, and obtaining an automatic measurement result of the railway gauge in combination with the local point cloud data. The problem that the railway gauge measurement result is inaccurate due to the fact that the railway section length is unreasonably arranged is solved.

The invention provides a two-stage pedestrian and vehicle detection method. The two-stage pedestrian and vehicle detection method specifically comprises the following steps: S1, collecting and marking a pedestrian and vehicle data set; s2, training a pedestrian and vehicle detection model; and S3, applying the pedestrian and vehicle detection equipment end. According to the method, the advantages of a single-stage model and a double-stage model can be effectively combined, some inherent problems existing in a single-stage detection model are optimized, through serial combination of the two single-stage detection models, the detection precision can be remarkably improved on the premise that the real-time performance is kept, false alarm and missing alarm are reduced, and meanwhile high robustness and flexibility are achieved. Compared with a single high-performance model, a two-stage detection model and a single lightweight model in the prior art, the method provides better balance in performance and resource occupation in a complex background, especially in a scene containing a small target, and is especially suitable for a scene needing to consider both speed and precision.

Embodiments of the present application relate to the technical field of human-computer interaction, and disclose a method and device for determining an interactive gesture and an electronic device. The method comprises the following steps: first, target component detection is performed; then, gesture detection and recognition are performed in a local area (around a boundary box) near the target component. In addition, for N gesture detection boxes (multiple users) in a current video frame, M target gesture detection boxes are selected therefrom to perform hand component positioning and gesture recognition processing, which can reduce detection time consumption and improve real-time performance. On this basis, the interactive gesture is determined according to the gesture detection result of the current video frame and the gesture detection results of a plurality of historical video frames, that is, gesture recognition is selectively performed on the user in each frame in a plurality of continuous video frames to determine the interactive gesture, which can reduce the computing power, reduce the time consumption, improve the real-time performance, and improve the accuracy and stability of the interactive gesture. That is, in the multi-user human-computer interaction scene, the interactive gesture can be accurately and timely determined.

The application discloses a kind of based on double filtering cooperation's remote sensing image scene classification method and system, in the method, for the problem that existing technology is difficult to give consideration to background noise suppression and key feature edge structure preservation when processing remote sensing image, a kind of double filtering cooperation optimization module is presented.The module in this paper introduces Gaussian filter smoothing channel to suppress unstructured high-frequency background noise, while introducing Gaussian Laplace edge enhancement channel to accurately capture and strengthen key geometric structure information such as feature contour;Through feature fusion, position coding, state space model and double activation gating mechanism, the multi-scale feature map is optimized and reconstructed to generate high signal-to-noise ratio and clear structure scene representation.The module is integrated into the remote sensing image scene classification model, and trained using focal loss, which can significantly improve the classification accuracy and robustness in complex texture interference scene.

The present invention provides a kind of low-temperature activated activated carbon, preparation method and application, which belong to the field of activated carbon nanomaterials. The raw materials used are common and easily available industrial-grade products, the equipment is simple, the preparation process is simple, the repeatability is strong, and it is conducive to industrial large-scale production. The present invention provides a kind of low-temperature activated activated carbon, the activation temperature is relatively low, and the corrosion of the activator to the equipment is relatively small, which not only further reduces the cost of the product, but also is human and environmentally friendly, and can be widely used in water treatment, chemical catalysts and carriers, solvent recovery, oil decolorization, air purification and other fields. Compared with the existing commercially available activated carbon, the product prepared by the present invention has a large specific surface area, a more developed pore structure, and a lower iron content in the product. In addition, the product of the present invention can be used in the fields of heavy metal ion adsorption in aqueous solution, automobile exhaust purification, impurity removal of printing ink, gas separation, supercapacitor electrode materials and other electrochemical energy storage materials.

The invention discloses an asphalt ultraviolet-visible absorption spectrum testing method based on an integrating sphere, belongs to the field of road petroleum asphalt, and solves the problem that asphalt cannot be uniformly mixed with powder to prepare a sample suitable for an integrating sphere method. The method comprises the following steps: adding an asphalt sample into an n-heptane-toluene mixed solution to form an asphalt dispersion solution; adding BaSO4 into the asphalt dispersion liquid, carrying out sealed standing, and then carrying out reduced pressure distillation to remove n-heptane and toluene so as to obtain BaSO4 powder with asphalt uniformly adsorbed on the surface as a to-be-detected sample; the method comprises the following steps: adding BaSO4 into an n-heptane-toluene mixed solution, sealing, standing, and carrying out reduced pressure distillation to remove n-heptane and toluene to obtain BaSO4 powder of which the surface does not adsorb asphalt as a blank sample; a to-be-tested sample and a blank sample are prepared into test pieces respectively, and the test pieces are tested by using an ultraviolet-visible spectrograph equipped with an integrating sphere. According to the method disclosed by the invention, the asphalt ultraviolet-visible spectrum which is more complete and accurate and has more obvious characteristics can be obtained.

The application provides a novel feature up-sampling method for semantic segmentation, which improves the resolution of an image by combining pixel reorganization Pixel Shuffle with an attention mechanism, prevents a chessboard effect, and avoids a 0 filling operation on unknown pixel points compared with a previous traditional up-sampling module; secondly, the attention mechanism is introduced to combine global information with local information, obtain a feature map with a larger receptive field and more detailed information, increase effective feature information, and realize the function of effectively improving the segmentation performance of the network. The application has universality, eliminates the limitations of previous methods by simply replacing all up-sampling modules, obtains more representative feature mappings, obtains an enlarged image with good features, makes the restored image obtain more feature information, and is beneficial to improving the segmentation result of the network model.

The invention relates to a method and a system for producing a relief on a substrate, the method comprising applying drops of a relief material (3) by digital inkjet printing (4), the relief material being deposited on a local area of ​​the surface of a substrate (1) to obtain a relief surface having a relief determined by the deposited relief material (3), a coating sheet (2) being bonded to the deposited relief material (3) and conforming to the relief surface, the coating sheet (2) being fixed by bonding by means of a glue (5) applied together with or in combination with the relief material (3).

The invention provides a method for selecting a reference adaptation area of an earth infrared image. The method comprises the following steps: establishing a global infrared radiation reference map and partitioning the global infrared radiation reference map; detecting a key angular point of each block region; a K-means algorithm is adopted to classify the key angular points, the classification number is determined by an improved weighted Voronoi graph, sub-regions are determined according to the classification result, it is ensured that the influences of all the sub-regions are similar, and therefore uniform distribution of features is achieved; constructing an infrared image suitability characteristic evaluation standard, counting infrared image suitability characteristic indexes of the sub-regions, and evaluating the infrared image suitability characteristic indexes; and based on the evaluation result of each sub-region, screening out a standard matching region meeting the requirement. According to the method, the infrared image reference matching area with remarkable features, rich information amount and high adaptability can be automatically generated. The technology provides solid data support for infrared image navigation, and has wide application prospects in the fields of image matching and satellite navigation.

The present invention relates to the technical field of image analysis systems, and in particular to a remote monitoring system for forest and grass coverage based on satellite remote sensing images, including a radar remote sensing sensor and a multispectral remote sensing sensor, and also including: an image acquisition module, an image analysis module, a key marking module, and a monitoring and control module. The present invention selects a standard optical monitoring image, that is, a standard optical image, according to the cloud cover, atmospheric light transmittance, and solar altitude angle at different optical monitoring image acquisition times, so that the standard optical image can be clear enough and the features of various objects in the image are obvious enough, thereby reducing the influence of weather changes and atmospheric cloud changes during the optical monitoring image acquisition process, and facilitating the use of the optical monitoring image as an analysis object to obtain an accurate and realistic forest and grass coverage.

The method and system for automatically extracting key frames of blast furnace material surface video provided by the present invention collects material surface video of blast furnace smelting process, locates the significant area of ​​material surface image based on boundary prior and image feature space, extracts accurate feature points of significant area of ​​image, and calculates pixel displacement of two adjacent frames of image according to optical flow method; then proposes a method of fusing local density maximum and GMM model clustering to identify material surface state; finally, divides the video collected by high-temperature industrial endoscope into different periods according to the change of density of image feature points, and extracts key frames of each material distribution period based on material surface state recognition result. The present invention can accurately identify material surface state and eliminate redundant information of material surface video, and automatically extract material surface video key frames with stable central airflow and obvious image features from blast furnace material surface video. At the same time, the present invention is suitable for automatic extraction of key frames of other videos with periodic changes and uneven image quality.

The invention discloses a novel agrobacterium tumefaciens-mediated injection method applied to instantaneous transformation of strawberry fruits, and belongs to the technical field of genetic engineering. By optimizing the agrobacterium injection method of strawberries and selecting injection varieties, the dyeing diffusion effect and the injection effect of injected fruits are remarkably improved, and the positive rate of instantaneous transgenic strawberry fruits is remarkably increased. Jiande red strawberries with a longer ginkgo period are selected as injection strawberry varieties, the characteristics are more obvious, and material selection is easy; according to the method disclosed by the invention, injection is carried out from the fruit stems of the strawberries, and genes carried by agrobacterium are injected into the central pith of the fruits through an injector, so that the genes are diffused in the whole fruits, the direction of vascular bundles of the strawberries cultivated under normal conditions is consistent, the diffusion of the injection is facilitated, and the method has important significance for enhancing the effect of instantaneous transformation of the strawberries.

The present application relates to a kind of urban green space scene laser radar point cloud single tree segmentation method, the method comprises: the point cloud of urban green space is carried out grid cubic processing;From the initial trunk position in grid is determined to fit DEM elevation plane, from DEM elevation plane upwards cutting tree bottom point to determine tree position;Using K-means clustering to carry out rough segmentation;Then according to single tree position, establish tree adjacent relationship, using ring extraction to carry out single tree optimization segmentation.The present application can make the trunk point cloud feature more obvious compared to tree crown point cloud, and the ground elevation information of the position where the trunk is located can be obtained, and compared with clustering time efficiency, it is based on the trunk point cloud and elevation information obtained to carry out a series of operations to complete rough single tree segmentation, and using adjacent relationship and ring extraction to optimize segmentation result, can make the accuracy of identification result higher.

The utility model discloses a functional all terrain vehicle high-safety integrated LED tail lamp. The functional all terrain vehicle high-safety integrated LED tail lamp comprises a tail lamp base, a tail lamp circuit board, a tail lamp middle cover and a tail lamp cover. The tail lamp base is fixed to a frame of the all terrain vehicle, a lamp panel mounting position is arranged on the tail lamp base, and the tail lamp circuit board is fixed to the lamp panel mounting position. The rear tail lamp middle cover is arranged on the rear tail lamp base in a buckling mode through a buckling structure and shields the rear tail lamp circuit board. The rear tail lamp lampshade is arranged on the rear tail lamp base through a buckling structure and shields the rear tail lamp middle cover, and the waterproof effect is achieved through the rear tail lamp lampshade. The sealing rubber ring between the tail lamp shade and the base is combined with the closed cavity formed by the buckling structure, so that water vapor and dust are effectively prevented from invading the internal circuit board, and the reliability and the long service life of the lamp under the severe working condition of the dune buggy are ensured.

The present invention relates to a toner. A toner comprising toner particles and an external additive A, wherein the toner particles contain a resin A, wherein the resin A is a resin represented by the following formula (1), the resin A is present on the surface of the toner particles, the external additive A is a fine particle containing silicon, the shape factor SF-1 of the external additive A has an average value of 105 to 120, the shape factor SF-2 of the external additive A has an average value of 100 to 130,

The invention discloses a multi-modal data generation method and device and a storage medium. The method and device are used for improving the overall training effect of a multi-modal model. Obtaining a multi-modal data set of the pre-trained multi-modal model; obtaining label information of the text data and the audio data according to the label information of the multi-modal data set; extracting a plurality of image data materials from the image data set according to the label information; inputting the image data material into a convolutional neural network model, and determining a plurality of feature regions on the image data material through a tag feature enhancement module according to the tag information; performing feature enhancement processing on each feature region according to hierarchical parameters among the tag information to generate tag features; generating a plurality of groups of normal distribution parameters for the label features through normal distribution sampling; and reconstructing the plurality of groups of normal distribution parameters and the corresponding label information through an encoder to generate multi-modal reconstructed image data.

A computer system for automated third-party cost estimation and production management receives a request for a specific packaging design template. The system utilizes a continuous monitoring software application to identify current system capabilities. The continuous monitoring software application is configured to query a plurality of packaging production machines located in different networks and different locations, collect current operating characteristics of each packaging production machine selected from the plurality of packaging production machines, and query queue storage space at each packaging production machine selected from the plurality of packaging production machines. Based on the current system capabilities, the system selects a specific packaging production machine from the plurality of packaging production machines to produce the specific packaging design template. The system then produces the specific packaging design template.

A foam-forming concentrate is provided. The foam-forming concentrate includes a surfactant component. The surfactant component comprises an anionic surfactant and an ampholytic surfactant. The foam-forming concentrate further includes a solvent component that includes a polar alkyl ether. The foam-forming concentrate further includes a gelling component selected from the group consisting of a carbohydrate, a synthetic hydrocolloid, or a combination thereof. The gelling component is substantially insoluble in the solvent component. The foam-forming concentrate has a viscosity of not greater than 2,000 centipoise. The foam-forming concentrate at a concentration of at least 1% (w / w) in water is capable of extinguishing fire fuelled by heptane at a flow rate of 3 gallons per minute for a period of less than 5 minutes, as determined according to UL 162.

The invention discloses a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation, and relates to the technical field of PCCP broken wire signal identification methods, and the method comprises the steps: S1, obtaining a broken wire signal, and analyzing the broken wire signal to obtain the frequency distribution feature of the broken wire signal, the method comprises the steps of S1, determining parameters of a band-pass filter according to frequency distribution characteristics, and setting the band-pass filter according to the parameters, S2, inputting a detection signal, and performing band-pass filtering processing on the detection signal through the set band-pass filter to obtain a filtered signal; s3, performing operation processing on the filtering signal to obtain a target modulus length; and S4, setting a target module length threshold value, and judging the occurrence moment of the PCCP broken wire signal. According to the method, the PCCP broken wire signal can be accurately identified, the operation efficiency is high, and the feasibility of real-time monitoring is improved.

The invention relates to a rolling bearing fault diagnosis method. The invention relates to the technical field of rolling bearing fault diagnosis, and the method comprises the steps: decomposing a bearing signal through peak envelope spectrum Fourier decomposition to obtain a component signal, carrying out the envelope spectrum analysis of the component signal, and converting a component envelope spectrum into a gray-scale map as an input feature of a convolutional neural network for fault diagnosis. According to the method, the original bearing signal is decomposed by using peak envelope spectrum Fourier decomposition, a frequency band with rich fault information is effectively separated from a noise frequency band, the component signal is converted from a time domain into an envelope spectrum and further converted into a feature enhanced grey-scale map, the obtained envelope spectrum grey-scale map has obvious fault features, and the fault detection accuracy is improved. And the problem that a traditional fault diagnosis method needs complex feature extraction is avoided. Diagnosis experiments are carried out on bearing vibration signals of different fault types, and the result shows that the method is higher in diagnosis precision and higher in anti-noise capacity.

The invention relates to the technical field of artificial intelligence, and particularly discloses a forged voice intelligent detection system based on multi-dimensional fusion, and the system comprises an acquisition module which is used for determining the importance score of target features, and screening to-be-selected features from the target features according to the importance score; the feature optimization module is used for establishing a forged voice detection model and determining a performance score of the forged voice detection model after training; sorting the performance scores, and determining a target performance score and a target set according to a performance score difference value; the feature selection module is used for determining a new training set and repeating the steps to obtain a sorting position interval and a target set; and iteration is carried out until the performance score meets a preset requirement, and standard features are determined and used for detecting the forged voice. According to the invention, the number of features required by forged voice detection can be reduced, and the detection efficiency is improved.

The invention relates to the technical field of image processing, in particular to a pavement pit detection method based on improved YOLOv5s, which comprises the following steps: sharpening an input image to make the pit edge more prominent and the pit feature more obvious, and improving the lightweight framework YOLOv5s and improving a backbone network in combination with a coordinate attention mechanism to make the pit edge more prominent and the pit feature more obvious. According to the method, the target feature extraction capability is enhanced, meanwhile, the detection capability is further improved by improving a frame loss function and replacing an original CIOU loss function with MPDIOU, and when the method is adopted for detecting the road surface pits, the generalization capability is higher, the recognition precision is higher, and the detection effect on small pits is higher compared with an original YOLOv5s method.

The invention relates to the technical field of data processing, and provides a rural cultural heritage protection method, which comprises the following steps: acquiring historical rural cultural heritage data, constructing a standard digital image, acquiring and processing the cultural heritage data in real time, performing feature extraction and coordinate data matching, comparing error information and determining an error region. And finally, performing field inspection by using inspection equipment. And when it is detected that the cultural heritage changes significantly, early warning information is generated in time. According to the method, the problem that the rural cultural heritage protection data cannot be dynamically detected in the prior art is solved, the protection efficiency and accuracy are improved, and powerful support is provided for sustainable development and inheritance of the rural cultural heritage.