1170results about How to "Fast extraction" patented technology

The invention discloses a method for recognizing a road traffic sign for an unmanned vehicle, comprising the following steps of: (1) changing the RGB (Red, Green and Blue) pixel value of an image to strengthen a traffic sign feature color region, and cutting the image by using a threshold; (2) carrying out edge detection and connection on a gray level image to reconstruct an interested region; (3) extracting a labeled graph of the interested region as a shape feature of the interested region, classifying the shape of the region by using a nearest neighbor classification method, and removing a non-traffic sign region; and (4) graying and normalizing the image of the interested region of the traffic sign, carrying out dual-tree complex wavelet transform on the image to form a feature vector of the image, reducing the dimension of the feature vector by using a two-dimension independent component analysis method, and sending the feature vector into a support vector machine of a radial basis function to judge the type of the traffic sign of the interested region. By using the method, various types of traffic signs in a running environment of the unmanned vehicle can be stably and efficiently detected and recognized.

Disclosed in the invention is a programmable visual chip-based visual image processing system, comprising an image sensor and a multilevel parallel digital processing circuit. The image sensor mainly includes a pixel array, an analog preprocessing circuit array and an analog-to-digital conversion circuit array; and the digital processing circuit consists of a parallel processing unit array with an M*M pixel level, a parallel processing unit array with M*1 rows, an on-chip artificial neural network and a reduced instruction processor dual-core subsystem. According to the provided system, high quality image collection with high speed and multilevel parallel image processing are realized and several high-speed intelligent visual application can be realized by programming; and compared with a traditional image system, the provided system has advantages of high speed, high integration, low power consumption and low cost. Moreover, the invention brings forward an embodiment for realizing the above-mentioned system as well as several high-speed intelligent visual image processing algorithms based on the embodiment. High-speed motion detection, high-speed gesture identification and rapid face detection are included; and the processing speed can reach 1000 frames per second. Therefore, a requirement of high-speed real-time processing can be met.

The invention discloses a fast image splicing method based on an improved SURF algorithm. An existing corner extraction method and an existing corner feature description method are improved, mismatching of extracted corners is eliminated, and multiple images can be spliced fast. At first, an improved FAST algorithm is adopted for increasing the extracted corners, the operation speed of the FAST algorithm for extracting the corners is high, and after the improvement is carried out, the stability is good; secondly, the combination of SURF description and LBP description is adopted for describing the corner feature, and in this way, the speed for matching the corners can also be increased; then an RANSAC method is adopted for eliminating mismatching, accuracy is improved, and a more accurate transformation matrix is obtained, so that fast splicing is carried out; finally, according to obtained matching point pairs, parameters for transforming images to be spliced into reference images are calculated, and a slow-in and slow-out method is adopted for finishing image splicing.

The invention discloses a real time panorama video splicing method based on ORB characteristics. The real time panorama video splicing method based on the ORB characteristics comprises the following steps: acquisition of multipath synchronized video data is started; pretreatment is carried out on images in various paths at a same moment, and color images are changed into gray scale images of 256 levels, and a de-noising processing is carried out on the images through employing a Gaussian filter; the ORB feature extraction algorithm is employed to carry out feature point extraction on the images in the various paths at the same moment, and ORB characteristic vectors of the feature points are calculated; through the adoption of the nearest neighborhood matching method and the RANSAC (random sample consensus) matching method to determine a homography matrix array between corresponding frames of the synchronized videos; frame scene splicing is carried out according to the homography matrix array; and finally spliced videos are output. The real time panorama video splicing method based on ORB characteristics and the apparatus are advantageous in that: the feature extraction speed and the coupling effect are improved in the image splicing process.

The invention discloses an ultrasonic phased array inspection oil-gas pipe circular weld defect automatic recognition method, in particular to an ultrasonic phased array inspection oil-gas pipe circular weld defect automatic recognition method for a support vector machine, which belongs to a pipe circular weld defect type recognition technology. The recognition method includes two parts of the construction of a support vector machine model and the automatic recognition of unknown defect types in circular weld. The construction steps of the support vector machine model include artificial defect signals extraction, wavelet decomposition promotion, feature extraction, feature optimization and the training of the support vector machine model. The automatic recognition steps of unknown defect types in circular weld include defect signals extraction, optimized feature calculation and automatic recognition based on the support vector machine. The invention combines a lifting wavelet transform with a fractal technology and has the advantages of high recognition accuracy and fast recognition speed. The automatic recognition of the defect types can be realized.

The invention relates to the field of molecular biology and discloses a kit and a method for extracting DNA from histiocytes. The kit comprises a binding solution, a regulating and binding solution and a deproteinization solution, wherein the binding solution comprises 3-5M of guanidinium isothiocyanate, 5-50mM of Tris-HCl with a pH value of 3.5-5.5, 5-25 percent by weight of Ttiton-X 100 and 5-20mM of Urea; the regulating and binding solution is isopropyl alcohol; and the deproteinization solution comprises 2-6M of guanidine hydrochloride, 5-50mM of Tris-HCl with a pH value of 6.5-8.0 and 5-50 percent by weight of absolute ethyl alcohol. The kit also comprises a pretreatment buffer solution, a splitting solution, a magnetic bead, a rinsing solution and an eluting solution. The DNA extracted by the kit and the method has the advantages of high content and high extraction speed, and the invention does not adopt toxic reagents and is suitable for extracting the DNA of various histiocytes.

The invention discloses a method for extracting circular facula sub-pel center and a device thereof. In the proposal of the invention, firstly, the initial position of the circular facula center is determined by utilizing a center of mass method; then, the pel position of the circular facula center is determined in a window by taking the initial position as a center according to the characteristic value of Hessian matrix; finally, image gradation function which is continuously distributed in a neighborhood of the pel position of the circular facula center is determined by using second order taylor expansion, and the position of extreme point of the image gradation function is the sub-pel position of the circular facula center. The searching range of the circular facula center is narrowed to a smaller area fast by the proposal of the invention, thus leading the follow-up operation such as gauss convolution, the calculation of S and the searching of the extremum to be carried out in the smaller area, and the calculated amount is small and the extracting speed is rapid. Therefore, the sub-pel position of the circular facula center is determined by the taylor expansion and the extracting precision of the circular facula center is ensured.

The invention relates to a Fe3O4 / C magnetic nano solid extracting agent coated with a novel carbon material, belonging to the field of chemical analysis and test instruments and equipment. The extracting agent simultaneously has large special surface area of nano material, short disperse path, magnetic separation capacity of magnetic material and strong adsorption capacity of carbon material, thereby being very suitable for pretreating large-scale and large-size water samples in batch. In addition, the solid extracting agent is simple to prepare, low in raw material cost, rapid and convenientto extract and friendly to the environment and only needs using a small amount of organic solvent in the whole process.

The invention relates to a method for extracting high-quality soybean germ oil by using subcritical butane and relates to the field of grain fine and deep processing. The method comprises: cutting purified soybean germs into slices which are 0.3 to 0.4 millimeters thick; placing the slices in an extraction kettle; performing 1 to 5 times of countercurrent leaching of the cut soybean germs by using liquefied butane as a solvent, wherein the extraction time is 10 to 120 minutes, the material to liquid ratio is 1:(1-6), the extraction temperature is 10 to 90 DEG C, and the extraction pressure is0.2 to 0.1MPa; delivering extracted mixed oil and wet dregs into an evaporation system and desolventizing under reduced pressure; and subjecting the obtained soybean germ crude oil to normal winterization, alkali refining and decolorization processes to obtain a soybean germ oil product with high beta-sitosterol content. In the invention, the soybean germ oil is extracted by using subcritical butane, the manufacturing cost of equipment is reduced considerably, the operation cost is low, the production period is shorter, the yield is high, and the method represents a low-consumption and environment-friendly practical technique for producing soybean germ oil.

The invention discloses a visual SLAM method based on point-line fusion, and the method comprises the steps: firstly inputting an image, predicting the pose of a camera, extracting a feature point ofthe image, and estimating and extracting a feature line through the time sequence information among a plurality of visual angles; and matching the feature points and the feature lines, tracking the features in front and back frames, establishing inter-frame association, optimizing the pose of the current frame, and optimizing the two-dimensional feature lines to improve the integrity of the feature lines; judging whether the current key frame is a key frame or not, if yes, adding the key frame into the map, updating three-dimensional points and lines in the map, performing joint optimization on the current key frame and the adjacent key frame, and optimizing the pose and three-dimensional characteristics of the camera;and removing a part of external points and redundant key frames; and finally, performing loopback detection on the key frame, if the current key frame and the previous frame are similar scenes, closing loopback, and performing global optimization once to eliminate accumulated errors. Under an SLAM system framework based on points and lines, the line extraction speed and the feature line integrity are improved by utilizing the sequential relationship of multiple view angle images, so that the pose precision and the map reconstruction effect are improved.

A method and device for controllably combusting combustible material, including a combustion device comprising an elongate combustion tube having an inlet section including an inlet for combustible material, an ignition section, including an igniter displaced along a length of the tube from the inlet section to ignite the combustible material, and at least one energy extraction device operatively coupled to the combustion tube and configured to extract energy from combustion of the combustible mixture.

The embodiment of the invention discloses a face recognition method and device as well as a mobile terminal. The method comprises the following steps of: pre-processing a target image, wherein the target image is to be judged whether to be a face image or not and the preprocessing step comprises geometrical normalization processing and illumination normalization processing; extracting face partial features by a partial filter set according to a Gabor algorithm; carrying out dimensionality reduction on the face partial features extracted by the partial filter set through a principal component analysis PCA and a linear distinguish analysis LDA; sending the face partial features subjected to the dimensionality reduction into a classifier to be matched and judging whether the target image is a face image or not. The face recognition method and device as well as the mobile terminal, provided by the embodiment of the invention, can focus on an embedded type application; the speed for extracting features is greatly improved, the storage amount is reduced, and the identification performance is better so that the face recognition method and device are suitable for an embedded type platform with a limited resource, such a mobile terminal and the like.

The invention relates to a robot forklift with a simple and convenient positioning function. The robot forklift is characterized by comprising a forklift body chassis, a driving wheel, follow-up wheels, forklift body stand columns, a forklift body control box, a positioning control device, a platform lifting device and a horizontal goods storing and taking device, wherein the driving wheel and the follow-up wheels are mounted under the forklift body chassis; the forklift body stand columns are mounted at side ends of the forklift body chassis and perpendicular to the forklift body chassis; the forklift body control box is mounted on a forklift body; the positioning control device comprises a positioning device for goods pallets on goods shelves, and a positioning device for discretionary object points on the ground. According to the robot forklift disclosed by the invention, the horizontal goods storing and taking device is mounted in lifting sliding chutes, and when the direction of the linear motion in a front-rear direction of the robot forklift is maintained to be constant, the horizontal movement of wheels is replaced by the horizontal movement of sliding plates, and the longitudinal positioning of the goods shelves is realized, so that not only is the horizontal deviation correcting problem solved, but also the longitudinal deviation correcting problem is also solved, and the extraction speed of goods is effectively increased.

The invention discloses an infrared image target recognition method. The method solves the technical problem of two-dimensional infrared image segmentation by using the concept of one-dimensional image processing, i.e. each row of image pixels for cyclic segmentation processing is synthesized into two-dimensional binary images; and the concrete steps of the method are as follows: firstly, the infrared image enhancement is performed by a multiscale wavelet transformation method; secondly, the stable segmentation of the infrared targets under various imaging conditions is realized by using the extraction to the convex area in the infrared image, and after each row of the pixels is processed, the processing results are synthesized into two-dimensional binary images with sizes consistent with that of the original images; thirdly, the binary image connected component labeling is performed to the segmented images, meanwhile, the basic characteristics are extracted, then the characteristics to be extracted are defined on the basis, and the characteristic extraction is calculated by using the basic characteristics; fourthly, a candidate target area is determined by a voting method, the accurate recognition is performed by ultimately combining the prior knowledge of the target to be recognized, and the final target is determined from the candidate targets. The invention has the advantages of accuracy, real-time and robustness.

The invention discloses a system for extracting a picture abstract based on page partitioning, which comprises a page preprocessing module, a page sorting module, a page partitioning module, a subject block identifying module and an information extracting module. The invention also discloses a method for extracting the picture abstract based on the page partitioning, which specifically comprises the following steps that: 1, a page is crawled on the Internet; 2, the page preprocessing module preprocesses the page; 3, the page sorting module sorts the preprocessed page; 4, the page partitioning module partitions a semantic chunk; 5, the page partitioning module sends the subject type page to the subject block identifying module, and the subject block identifying module identifies the subject block and then sends the subject block to the information extracting module; and the page partitioning module sends the non-subject type page to the information extracting module; and 6, the information extracting module downloads the picture and is correlated with the page. The system and the method have the advantages of high extracting speed, high accuracy and good effect.

The invention discloses a pedestrian image detection method based on sparse coding and a neural network. The method comprises the steps: preprocessing the samples in a sample data set, and extracting the features of an aggregation channel; constructing a BP-AdaBoost strong classifier model and training; preprocessing video frames, and scaling the video frames with different scales to obtain an image pyramid; utilizing the quick image feature pyramid calculating method to integrate with a sliding window detection method to calculate the window images to obtain the features of the aggregation channel, and utilizing the trained strong classifier to perform classification; and when the detection result of the window images is a positive sample, outputting the detection window, and fusing all the detection windows to obtain an accurate positioning window. The pedestrian image detection method based on sparse coding and a neural network is high in the calculating speed and the accuracy when the training samples are large.

The invention discloses a method for rapidly detecting polybrominated diphenyl ether residue in a sample, particularly in textiles. The method is characterized by comprising the following steps: (1) the sample is crushed, dipped in base solution and ultrasonically extracted in water bath, and solid phase microextraction is employed to concentrate a target compound; and (2) polybrominated diphenyl ether is qualitatively and quantitatively determined by gas chromatography-mass spectrometry after thermal desorption. The method helps effectively solve the problem of matrix effect that fabric fibers have no solvent extraction, obtains a satisfactory result for the analysis of a selected target compound, has the advantages of low organic solvent consumption, rapid extraction speed, less interference from impurities, high sample recovery rate and the like, and can be applicable to rapidly detecting polybrominated diphenyl ether flame retardant residue in textiles.

The invention provides a magnetic mesoporous silica gel material extracting agent used for enriching and / or separating trace organic pollutants in an environment water sample, and a preparation method thereof. A core of the extracting agent provided by the invention is a silicon gel microsphere (i.e. magnetic silicon gel microsphere) wrapped with a plurality of Fe3O4 magnetic nano particles, and a shell of the extracting agent is a mesoporous silicon gel shell layer with inner and outer surfaces modified by C18 silanization. The extracting agent integrates huge special surface area, excellent extracting capacity and ultrahigh extracting volume of a mesoporous material, magnetic separating capacity of a magnetic material, strong extracting capacity of C18 groups and special volume exclusion function of a vertical channel of the mesoporous material. In addition, the extracting agent provided by the invention has the advantages of high extracting efficiency, low preparation cost and simple preparation method, and is very suitable for preprocessing samples of the trace organic pollutants in a bulk-mass water sample.

The embodiment of the invention discloses a key frame extraction method and device and a storage mediumThe embodiment of the invention includes: acquiring the video frame set corresponding to the video,wherein the video frame set comprises a plurality of video frames; determining a current reference video frame in the video frame set, extracting a corresponding video frame from the video frame setaccording to the reference video frame to serve as a target video frame, acquiring similarity information between the target video frame and the reference video frame, and determining the target video frame as a key frame when the similarity information meets a preset condition. According to the scheme, the video key frames can be extracted based on the similarity between the video frames;, the effective video key frame can be quickly extracted from the video, the speed of extracting the video key frame is increased, the scheme does not depend on the frame rate of the video, the scheme can beapplied to videos of various frame rates, and the accuracy and flexibility of extracting the video key frame are improved.

The invention discloses a preparation process for extracting active ingredient of natural plant by supersonic, which comprises the following steps: firstly, cleaning and removing dust; secondly, crushing; thirdly, steeping: placing the crushed natural plant raw material to be extracted into a material mixing tank with a mixer for continuously mixing, adding steeping fluid in the natural plant rawmaterial to be extracted in the material mixing tank and obtaining material mixture after mixing and stirring; fourthly, supersonically extracting: passing the material mixture circularly continuously through the inner part of a supersonic extraction device by a circulation pipe between the material mixing tank and the supersonic extraction device and performing the continuous, circulated and supersonic extraction; fifthly, filtering, sixthly, reducing and concentrating. The invention saves time and labor, can obviously improve the product quality and the product yield, has short extraction time, high extraction efficiency, and low temperature during the extraction without heating, and prevents effective ingredient being damaged when heating at high temperature.

The invention relates to a solid phase extraction column based on graphene bonded silica gel, which is used in concentration and separation of organic pollutants in an environmental water sample. A graphene bonded silica gel padding is prepared by taking 3-aminopropyl triethoxy silane (APTES) as the cross-linking agent; then graphene is bonded on the surface of silica gel; after product graphene bonded silica gel granules are evenly filled in a solid phase extraction column, the graphene bonded silica gel based solid phase extraction column is obtained. The performance indicators of the solid phase extraction column are tested through polycyclic aromatic hydrocarbons, including naphthalene, acenaphthylene, fluorene, phenanthrene, fluoranthene and pyrene, and the test result shows that the solid phase extraction column has good extraction efficiency, reproducibility and stability. The graphene bonded silica gel has the characteristics of big specific surface area, strong Pi electron adsorption ability and high extraction capacity of the graphene, and the stable mechanical property and good adsorption ability of the silica gel. Moreover, the solid phase extraction column based on the graphene bonded silica gel is low in preparation cost and simple to operate and is suitable to be used for pretreatment of low-concentration organic pollutants in a big-volume water sample. The solid phase extraction column has a good application prospect in the concentration and separation of environmental organic pollutants.

The invention relates to a non-contact vibration detecting method of a tower crane structure and belongs to the technical field of detecting. The method is non-contact measurement. A multipoint and on-line test can be achieved by arranging a plurality of cameras without arranging a sensor on a detected object. The non-contact vibration detecting method of the tower crane structure has no load effect and is easy to implement. Image information is directly digitalized to form a full-automation and full-digitalization vibration measurement disposing and recording system. Amount of information is larger than that of a traditional method. Regions of interest (ROI)technology is used in the method to reduce an image tracking area and calculated amount. Speeded-up robust features (SURF) method is used to accelerate extraction rate of feature points. The PauTa criterion is used to reject'outlier' and improve accuracy of the SURF method. Reliability of a crane safety monitoring system in prior art is improved.

The invention discloses a preparation method of a cannabinol compound. The method comprises the following steps: screening industrial cannabis plant materials; drying and carrying out ultra-micro pulverization; putting the materials into an ultrasonic extraction device with a stirring function; adding a weak-polar or non-polar organic solvent and weak acid, and carrying out ultrasonic assisted extraction; filtering and adding active carbon in filtrate; filtering and evaporating to remove the solvent so as to obtain extractum of industrial cannabis extract; sufficiently dissolving the extractum by adopting a high-polar organic solvent; freeing, sealing, filtering and evaporating to remove the solvent, so as to obtain a primary product of the industrial cannabis extract; stirring the primary product and slowly adding an alkali solution; centrifuging and filtering; slowly adding an acid solution into filtrate; separating to obtain an industrial cannabis extract product containing the rich cannabinol compound. The method disclosed by the invention has the advantages of high extraction efficiency, low cost, few product impurities, small pollution and easiness of being industrially realized.

This invention provides a method for separating and extracting high-purity stachyose product from Rehmannia glutinosa and Stachys sieboldii. This invention also provides a method for continuously separating and extracting stachyose and catalpol from Rehmannia glutinosa. The method comprises: separating and purifying the extract by passing through continuous pressure-filtration, microfiltration, ultrafine filtration and nanofiltration membranes, and active carbon column, macroporous adsorption resin column, cation exchange resin column, and anion exchange resin column chromatography to obtain stachyose and catalpol. The method has scuh advantages as increased product quality, low wnwegy consumption, and increased raw material utility.

A mold monitoring method based on a FAST-9 image characteristic rapid registration algorithm include the following process: when mold open of an injection molding machine is in place and after an ejector pin is ejected, standard form images are acquired respectively; operating state information of the injection molding machine is waited, when the injection molding machine is operated to the place of the mold open, images of surfaces of a mold cavity are taken continuously through a camera, current frames of the monitored images are preprocessed, and preparations are made for subsequent image rapid registration; the FAST-9 image characteristic rapid registration algorithm is carried out; after registration, differences are made between the current frames and the form images; through adoption of the ostu automatic threshold value partitioning algorithm, binaryzation of the images is achieved, and continuous closing and opening operation is conducted on the images; whether anomaly exists in molding of products is checked through electron hole detection, when the anomaly exists, alarm information is displayed; otherwise, the operating state information of the injection molding machine in the next period is waited. The mold monitoring method based on the FAST-9 image characteristic rapid registration algorithm is good in real-time performance and strong in robustness.

The invention relates to video copy detection system and method, which are used for fast and accurately checking that whether input copies a video segment in a video data set and outputting a beginning position and an end position in the presence of copied segment according to a query video input by users. The method comprises three steps of feature extracting, feature matching and amalgamation judging. The SURF (Speeded Up Robust Features) feature of a video frame is firstly extracted, an optimization scheme of an integrogram is utilized in the extraction process of the feature, and the extraction speed is high. The feature matching step is different from traditional methods for matching feature vectors of each feature point, and adopts a two-layer matching method which comprises the following steps of: firstly, adopting a bag-of-words method on the feature vectors of each key frame, obtaining a word frequency histogram of the key frame, and then indexing the word frequency histogram of each key frame for researching a matched key frame pair; and finally matching the feature points in the key frame pair. In the amalgamation judging step, a probabilistic graphical model is established for PSE (Product Safety of Electrical Appliance and Materials), a powerful reasoning method is utilized for deducing the existence and the position of the copied segment, fully the time consistency and the space consistency of the video are fully utilized, and the disadvantages of traditional amalgamation methods are avoided.

The invention discloses a method for automatically extracting information of a bare area in a slumped mass. The method comprises the following steps of: (1) calculating the vegetation index and the soil brightness index of ground features in a remote sensing image, and respectively generating a vegetation index grayscale image and a soil brightness index grayscale image; (2) transforming principal components of the remote sensing image and extracting a first principal component to generate a first principal component grayscale image; (3) synthesizing the vegetation index grayscale image, the soil brightness grayscale image and the first principal component grayscale image into a colorful synthetic image; (4) transforming the principal components of the colorful synthetic image and extracting a first principal component to generate a new first principal component grayscale image; and (5) carrying out threshold segmentation on the new first principal component grayscale image to obtain a binary image of the bare area in the slumped mass, thus the information of the bare area in the slumped mass is extracted. The method disclosed by the invention can be used for accurately extractingthe information of the bare area in the slumped mass; the extracting accuracy and the extracting speed are high; and the influence from topographic shadows, vegetation and other ground features can be eliminated effectively.

The invention relates to a keyword extraction method and system for judicial text data, and the method comprises the steps: firstly, introducing a legal name and judicial related professional vocabulary for word segmentation, and carrying out the manual reinspection to construct a judicial professional vocabulary annotation table; secondly, by constructing a judicial professional vocabulary labeling dictionary and a large-scale user dictionary, carrying out word segmentation, removing stop words and the like, and obtaining words; collecting and counting keyword search vocabularies of various disputes and causes to form candidate keywords; then, adding a title word weight and word global weight value method to correct the weight of the candidate keyword TF _ IDF; and if the to-be-extracteddocument does not contain the candidate keywords, inputting the TF _ IDF normalization value of each word in the document as an initial weight of a TextRank algorithm to obtain a final word weight. According to the method, judicial text data can be well matched, matching performance is high, and the method is suitable for most judicial text data. Moreover, extraction speed is increased, and extraction accuracy is high.

The invention relates to a linked parasitic capacitance extracting method, which can automatically adapt characteristic size, belonging to the integrated circuit computer aided design, wherein arrange the scan band width reduced in ratio with the characteristic size to save memory; reduce the environment conductor horizontally according to the characteristic size after cutting to improve the capacitance extracting speed, and increase the environment conductor longitudinally to improve the extracting accuracy; then divide the edge length from the relative length of characteristic size, to be divided with edge unit to improve the extracting accuracy; at last, solve the discrete edge integral equation to obtain the parasitic capacitance between conductors to improve the property of linked parasitic capacitance extracting program.

The invention discloses a palm detecting and key point location method based on deep learning. The palm detecting and key point location method based on deep learning particularly includes the following steps of 1, collecting training samples; 2, building network models, wherein a CNN feature extraction network, an RPN candidate area extraction network and a judgment network are built; 3, trainingthe network models, wherein the CNN feature extraction network, the RPN candidate area extraction network and the judgment network are initialized; 4, building a detection model, wherein the CNN feature extraction network, the RPN candidate area extraction network and the judgment network form a Faster R-CNN detection network; 5, detecting the palm and location key points. According to the palm detecting and key point location method based on deep learning, the Faster R-CNN detection framework which has optimal performance and accuracy at present is adopted, compared with a Fast R-CNN, a RPNis adopted to replace a Selective Search method to extract candidate areas, the RPN is completely built in the whole target detection framework, therefore, the speed of extracting the candidate areasis increased, and at the same time, the detection accuracy is improved.