290results about How to "The detection process is fast" patented technology

The invention discloses a brake spring piece flatness detection system and method, and relates to the technical field of automobile part processing. The brake spring piece flatness detection system comprises a data acquisition module, a data analysis module, a data screening module, a dynamic monitoring module and a comprehensive judgment module, wherein the data acquisition module is used for acquiring static detection data of N groups of spring pieces randomly extracted from the same batch; the static detection data comprises top view shooting picture data and side view shooting picture data of N groups of spring pieces, which are acquired at equal intervals by utilizing an optical detection instrument, and size data of the spring pieces, which are acquired by utilizing a laser radar, and the technical key points are as follows: through analysis and processing of the static detection data and the dynamic detection data, the overall quality of the spring pieces is shown; related workers can conveniently and quickly know whether the spring pieces meet the requirements of automobile braking or not, high-quality products can be quickly screened out from numerous spring pieces, use is convenient, secondary detection is not needed in the follow-up process, and good application prospects are achieved.

The invention provides a hyperspectral image classification method based on a state space model. The problem that an existing method is poor in classification precision is mainly solved. According to the scheme, the method comprises the following steps: 1) inputting a hyperspectral image, and preprocessing the hyperspectral image; 2) eliminating spectrum variability by using a neighborhood spectrum fusion method, sensing spectrum detail differences by using a sub-spectrum scanning method, and obtaining image spectrum classification information; 3) constructing a spatial feature extraction network by using an SS2D layer, an improved spatial attention mechanism and a plurality of residual blocks, and acquiring spatial classification information of the hyperspectral image by using the network; and 4) fusing space and spectrum classification information at a decision layer to obtain a final classification result. According to the method, the influence caused by spectrum variability can be reduced in a relatively small range, local spectrum information is not ignored on the premise that comprehensive spectrum information can be perceived, and spatial information is combined, so that the classification effect of the hyperspectral image is effectively improved.

The invention provides a traveling wave head extraction method and system, electronic equipment and a storage medium. The method comprises the following steps: detecting an original traveling wave signal to obtain the original traveling wave signal; calculating the fast Fourier transform of the original traveling wave signal and the fast Fourier transform of the window function introducing the correction factor so as to respectively obtain a discrete form of the original traveling wave signal and a Gaussian function under the optimal frequency; calculating a product of a discrete form of the original traveling wave signal and a Gaussian function under the optimal frequency according to the frequency sampling point, and calculating to obtain discrete S transformation; intercepting two-dimensional diagrams of the traveling wave fault signal with time and amplitude as an x axis and a y axis; on the basis of discrete S transformation of the traveling wave, the amplitude difference of adjacent time nodes in the two-dimensional diagram is adopted, a threshold value is set at the same time, the detection time of a first traveling wave head is obtained, the position of a sudden change point is determined, and the traveling wave head is extracted. According to the method, various fault transient states can be detected, the detection speed is high, the sensitivity is enough, and the influence of noise can be effectively reduced.

The invention relates to the technical field of mark point identification and positioning, in particular to a polygon mark point rapid identification and positioning method and system and a storage medium. Then, the gradient direction information and the coordinate information of the reserved strong edge are used for rapidly estimating the rotation angle and the center of the polygon mark point to be detected, the method is small in calculation amount and easy to operate, and rough pose information of the polygon mark point to be detected can be rapidly obtained; the estimated angle is further corrected in a small range in a rotary projection mode, the deviation between an estimated value and a real value is reduced, and a more accurate initial value is provided for subsequent registration; finally, the corrected coarse positioning result is used as an initial value for registration, the convergence speed can be increased, the number of iterations can be reduced, meanwhile, the probability that iteration falls into a local extreme value can be reduced, the operation speed is increased, and the accuracy rate is increased.

The invention discloses a kit for detecting salmonella enteritidis and application of the kit, and belongs to the technical field of biomolecular detection. The kit comprises the following components: an allosteric aptamer probe of which the nucleotide sequence is as shown in SEQ ID NO.1, a template chain of which the nucleotide sequence is as shown in SEQ ID NO.3 and a fluorescence report probe of which the nucleotide sequence is as shown in SEQ ID NO.4. According to the kit, salmonella enteritidis can be distinguished from other impurities, and the specificity is high; a visual fluorescence signal can be obtained and is easy to interpret; according to the invention, a template chain can be cyclically utilized in the reaction to generate a Trigger chain to guide the TtAgo protein to efficiently and specifically cut a fluorescent reporter probe, the output signal is increased, the detection time is short, the detection limit is low, the sensitivity is high, the specificity is good, and the method has practical popularization and application prospects.

The invention discloses a catenary dropper unsupervised defect detection method based on a convolution automatic encoder, and aims to solve the problems of low efficiency, high cost, high false detection rate and high omission ratio in a traditional dropper detection method based on manual inspection or simple image processing. The method comprises the following steps that firstly, image data of the catenary dropper is collected and preprocessed, and the preprocessing step comprises image graying, denoising and normalization processing so as to eliminate the influence of external factors and enhance the image quality; then, the preprocessed image is input into a convolution automatic encoder model, a convolution automatic encoder is composed of an encoder and a decoder, the encoder extracts local features of the image through convolution and pooling operation and carries out dimension reduction, potential space representation is generated, and the potential space representation is extracted; the decoder restores the potential spatial representation to an output image similar to the input image by a deconvolution or up-sampling operation. By calculating a reconstruction error between an input image and a reconstruction image, an abnormally large reconstruction error is identified so as to judge whether the dropper has defects or not.

The invention discloses a method for detecting position information of a butt joint hole of a skirt end face of a solid rocket engine, and the method comprises the steps: 1, installing a tool on the butt joint hole of the skirt end face, and completing the laser scanning of the skirt end face and a cylinder segment which is about 200mm around the skirt; the tool is conical, the tail end of the tool is provided with threads, the tool can be directly installed on a skirt end face threaded hole, and the axis of the conical face is coaxial with the axis of the threaded hole. 2, performing preliminary optimization on the three-dimensional point cloud model obtained by scanning, then eliminating errors caused by noisy points by using a filtering model combining a system filtering algorithm and a moving least square method, and finally completing overall optimization of the point cloud data; 3, packaging the optimized model by using a triangle principle, carrying out reverse modeling on the packaged three-dimensional model, and extracting to-be-detected data by using the reverse regularized model and the three-dimensional point cloud model; according to the invention, detection of position information of butt joint holes of skirt end surfaces of various models of solid rocket engines including large diameters can be realized.

The invention discloses an SERS (Surface Enhanced Raman Scattering) sensor for detecting diabetic retinopathy serum biomarkers miR-21 and miR-152 as well as a preparation method and application of the SERS sensor, and belongs to the technical field of biomedicine. The preparation method comprises the following steps: self-assembling AuNHs through an oil-water interface to obtain an AuNHs array, and modifying an aptamer chain cDNA on the surface of the array to obtain an SERS (Surface Enhanced Raman Scattering) substrate; the method comprises the following steps: modifying a Raman signal molecule 4-MBA and a nucleic acid aptamer complementary chain H1 to the surface of AuNTs to obtain an SERS probe 1; a Raman signal molecule DTNB and a nucleic acid aptamer complementary chain H2 are modified to the surface of AuNTs, and an SERS probe 2 is obtained; and combining the prepared SERS probe with a complementary basic group in the nucleic acid aptamer chain cDNA on the SERS substrate, and fixing on the AuNHs array to construct the SERS sensor.

The invention discloses a composite material plate ultrasonic detection robot and a detection method.The detection robot comprises a driving module, a guide rail support module and a main control module, rubber suction cups are arranged at the bottom of the driving module and are arranged at the two ends of the guide rail support module, and the main control module comprises a phased array module, a phased array probe and an electric telescopic rod; the driving module is arranged on the cylindrical support and moves left and right along the cylindrical support, the driving module establishes connection with the main control module through the wireless transmission module, and the main control module transmits instructions and data through the wireless transmission module; during detection, the rubber suction cup of the detection robot is firmly adsorbed on the surface of a detected composite material plate, and the phased array module collects detection data in real time and transmits the detection data to the mobile terminal equipment. The device is used for ultrasonic nondestructive testing, modular design and wireless control of non-metal composite material large plates, can quickly and efficiently realize full-automatic ultrasonic testing, displays the quality condition of the composite plates in real time, and is high in detection speed and high in detection precision.

The invention belongs to the technical field of varnished wire detection, and particularly relates to a detection device for an inductance element varnished wire, which comprises a base, a winding traction mechanism, a guide diameter measuring mechanism, a pressurization air supply mechanism, a wire diameter feedback regulation and control mechanism, a detection angular velocity control mechanism, a detection result visual display mechanism and a PLC (Programmable Logic Controller). According to the invention, continuous and relatively stable tensile force is provided through constant-speed rotation of the winding drum, the increase of the tensile force is more linear, additional damage to the enameled wire caused by sudden impact force is avoided, the test result can more accurately reflect the tensile property of the enameled wire in actual use, and the test can be carried out in a relatively compact space, so that the test efficiency is improved. According to the device, the requirement for a site is relatively low, the angular speed of the winding drum and the detection height of the winding drum in the detection process can be automatically adjusted, and a final test value in a tension test can be displayed more visually and clearly.

The invention discloses a focus detection method and device based on deep learning, and relates to the technical field of deep learning, and the method comprises the steps: obtaining a capsule endoscopy image sequence, and carrying out the video frame extraction of the capsule endoscopy image sequence, so as to obtain a gall bladder endoscopy image; constructing a focus detection neural network for detecting the focus of the gallbladder endoscopic image, and performing multi-level feature extraction and multi-level feature fusion on the focus of the gallbladder endoscopic image based on the focus detection neural network to obtain a preliminary detection result; and constructing a deep calculation neural network for deeply extracting focus features and calculating the number of the focus, performing deep fusion feature extraction on the gallbladder endoscopic image according to the preliminary detection result based on the deep calculation neural network, refining repeated judgment of the focus, and generating a final detection result of the focus. And counting and analyzing the lesions according to the final detection result to obtain the number of the lesions and lesion information of each lesion. According to the invention, focus detection speed and accuracy are improved.

The invention relates to the technical field of molecular biological detection, and particularly discloses a quadruple fluorescent PCR primer probe combination, a reaction system and a detection method for identifying a mycoplasma gallisepticum wild strain and a vaccine strain. The quadruple fluorescent PCR primer probe combination for identifying the mycoplasma gallisepticum wild strain and the vaccine strain comprises a universal primer probe group for detecting mycoplasma gallisepticum and a specific primer probe group specially aiming at three vaccine strains F-36, TS-11 and 6 / 85 respectively. According to the quadruple fluorescent PCR primer probe combination, the reaction system and the method for identifying the mycoplasma gallisepticum wild strain and the vaccine strain, the purpose of simultaneously detecting and distinguishing DNA of the mycoplasma gallisepticum wild strain and DNA of three different vaccine strains (F-36, TS-11 and 6 / 85) of the mycoplasma gallisepticum wild strain can be achieved, so that detection and differential diagnosis are more efficient, economical, simple and convenient; the kit has the advantages of high sensitivity, good specificity, rapidness, high flux, simplicity and convenience in operation and the like.

The present invention provides an eccentric shaft inspection tool for realizing the inspection of the eccentric shaft, wherein the eccentric shaft comprises a first eccentric shaft, a first cam, a second eccentric shaft and a second cam. The inspection tool comprises a base, a first cam sleeve, a first eccentric shaft sleeve, an integrated sleeve, a measuring bar and a height gauge. A first shaft hole matching the first cam is formed on the first cam sleeve; a second shaft hole matching the first eccentric shaft is formed on the first eccentric shaft sleeve, and a predetermined first eccentricity is provided between the second shaft hole and the first shaft hole; a third shaft hole and a fourth shaft hole are formed in the integrated sleeve, the third shaft hole matches the second cam, the fourth shaft hole matches the second eccentric shaft, and a predetermined second eccentricity is provided between the third shaft hole and the fourth shaft hole; the first end of the measuring bar is connected to the integrated sleeve, and the second end is a free end. The present invention has a simple structure, a fast inspection speed and a high inspection accuracy, and it improves inspection efficiency and reduces inspection costs.

The invention relates to the technical field of machine vision, in particular to a visual box grabbing method for a gantry crane, which comprises the following steps: mounting four groups of high-definition cameras on the same plane of the lifting end of a lifting appliance, and enabling the four groups of cameras to acquire images in real time in up-down and left-right symmetry; based on images shot by a high-precision calibration plate in an operation area, internal parameters and external parameters of four groups of cameras are calculated by applying a calibration algorithm; analyzing container corner image features, and determining corner three-dimensional coordinates and image pixel coordinates based on a measurement tool; a multi-task learning model is established, angular point positioning is used as a main task, the height change of the lifting appliance is simulated by cutting pictures, data enhancement training is carried out, and the method has the advantages of being easy to install, high in detection speed, high in precision, high in robustness and the like.

The invention belongs to the field of medicine crystal form quality control, and particularly relates to a quantitative method for a lamidipine hemisuccinate dihydrate crystal form D. The determination method has high specificity and accuracy for determination of the content of the lamidipine hemisuccinate dihydrate crystal form D in the lamidipine tablet. The content of the low-concentration crystal form D can be accurately quantified, and a research thought and a feasible method are provided for detecting the low-content crystal form in a medicinal preparation.

The invention relates to the field of environmental chemistry and biological sample analysis, in particular to an analysis method for detecting oxidative stress biomarkers in urine. The analysis method capable of simultaneously detecting dityrosine, 8-hydroxyguanosine, 8-hydroxydeoxyguanosine and 4-hydroxynonenal mercaptouric acid in human urine is provided on the basis of liquid chromatography-tandem mass spectrometry, effective purification of a urine sample is achieved through specific pretreatment, the matrix effect is reduced, and the detection sensitivity is improved. And specific liquid chromatogram parameter conditions are combined, so that simultaneous detection of the four oxidative stress biomarkers is realized. The method is good in selectivity, high in sensitivity and high in detection speed, the problem that polar and non-polar components cannot be measured at the same time at present is solved, the contained oxidative stress biomarkers are more comprehensive and more reasonable, and the oxidative stress damage condition of the organism can be better reflected.

The invention provides a real-time safe distance monitoring method and system for field operation and a storage medium, and the method comprises the following steps: calculating the distance between a field operator and a potential dangerous object or region as a first distance based on a real-time image of the field operation, and executing the subsequent steps if the first distance is smaller than a first set value; otherwise, continuing to calculate the first distance in real time; based on the real-time three-dimensional position data of the field operation, calculating the distance between the field operation personnel and the potential dangerous object or area as a second distance; performing data fusion on the first distance and the second distance to obtain a third distance; if the third distance is smaller than the second set value, alarm operation is executed; otherwise, continuing to calculate the first distance in real time. According to the invention, the technical advantages and advancement of various intelligent detection methods are fully combined, and the safety distance of field operation is effectively monitored in real time.

The invention discloses a network log anomaly detection method and system based on GPT2-PCA. The method comprises the following steps: collecting a log message; preprocessing the log messages to obtain a log message set; performing structured representation on the log message to obtain a structured log set; serializing to obtain a serialized log set; pre-training a GPT-2 model through the serialized log set, and performing feature extraction on log messages by using the trained GPT-2 model to obtain a log embedding vector set; and training a PCA-based classification model according to the log embedding vector set, and carrying out anomaly detection by using the trained PCA-based classification model. According to the method, fine-grained preprocessing of logs of various devices and applications can be supported, the normal behavior mode of the logs can be captured in a longer context, and the detection speed and precision can be effectively improved.

The utility model discloses an automatic packaging detection circuit for a tantalum capacitor, which comprises a square-wave generator, a current reversing switch, a first detection branch and a second detection branch, and is characterized in that the square-wave generator and the current reversing switch are connected in series and then are sequentially connected with a resistor Rcs and a detected capacitor CDUT; the first detection branch and the second detection branch are connected in parallel; the square wave generator is used for charging the tested capacitor CDUT; the current reversing switch is used for controlling the charging and discharging current of the capacitor CDUT; the first detection branch is used for carrying out open circuit detection on the capacitor CDUT; and the second detection branch is used for carrying out short circuit detection on the detected capacitor CDUT. According to the utility model, the tantalum capacitor packaging is detected by adopting the circuit, the detection circuit is simple in structure, low in cost, high in detection speed and very suitable for automatic detection equipment, multiple groups of circuits can be tested at the same time, multiple devices can be tested at a time, and the test efficiency is greatly improved.

The invention relates to the technical field of sensor preparation and antibiotic detection, in particular to a preparation method of a dual-emission fluorescence sensor, a product and application of the dual-emission fluorescence sensor. The preparation method of the dual-emission fluorescence sensor comprises the following steps: dissolving 1, 3, 5-benzene tricarboxylic acid and melamine in an organic solvent to prepare a suspension; the suspension liquid is subjected to a first heating reaction and then mixed with zinc salt to be subjected to a second heating reaction, and Zn-MBTC is obtained; the preparation method comprises the following steps: dissolving Zn-MBTC and Ru (bpy) 3 < 2 + > in water, carrying out ultrasonic treatment and oscillation, and then washing, centrifuging and drying to obtain the dual-emission fluorescence sensor. The dual-emission fluorescence sensor prepared by the method has the advantages of high sensitivity and high detection speed when being used for detecting four tetracycline antibiotics, namely doxycycline, tetracycline, oxytetracycline and gold tetracycline, and the detection limit is as low as 0.013 mu mol / L.

The invention discloses a packing material welding quality detection device which comprises a first detection electrode and a second detection electrode which are arranged in a spaced mode, a detection area is formed between the first detection electrode and the second detection electrode, and the first detection electrode is provided with at least two first detection parts capable of controlling power-on and power-off respectively. The second detection electrode is provided with at least two second detection parts capable of respectively controlling power-on and power-off, and orthographic projections of the first detection parts and the second detection parts in the detection area have overlapped parts. The detection method comprises the following steps: S1, placing a welding area of a packing material in a detection area between the first detection electrode and the second detection electrode; s2, power-on and power-off of the first detection parts and the second detection parts are controlled correspondingly, and the welding quality of the corresponding positions of the overlapping parts is detected one by one; if the detection capacitors corresponding to all the overlapping parts are located in the set interval, it is indicated that the welding quality is qualified, and if the detection capacitor of any overlapping part is not located in the set interval, it is indicated that the welding quality is unqualified, and detection is directly ended.

The invention relates to the field of microbiological detection, in particular to a primer combination, a kit and a method for detecting balloonium martensii. The primer combination comprises an upstream primer and a downstream primer, the nucleotide sequence of the upstream primer is as shown in SEQ ID No.2, and the nucleotide sequence of the downstream primer is as shown in SEQ ID No.3. By utilizing the primer combination, the kit and the PCR detection method provided by the invention, the balloonium botryoides and the balloonium viride can be distinguished, and the PCR detection method provided by the invention can be used for efficiently, rapidly and specifically detecting the balloonium botryoides.

The present invention relates to a rapid detection method and device for soil organic matter based on near-infrared spectroscopy technology. Soil spectrum information at different humidity levels is collected by a self-made near-infrared spectroscopy detection device. The spectrum contains information about soil organic matter. The soil moisture content of 12% is used as the reference humidity. The spectrum is corrected by a polynomial smoothing algorithm and a humidity self-correction algorithm to eliminate the influence of different soil particle sizes and different moisture contents. Then, the physical and chemical true value of organic matter in the soil is determined by an acidic potassium dichromate volumetric method, and a one-to-one correspondence between the corrected spectral information and the organic matter content is established. The characteristic wavelength is then selected by combining a competitive adaptive reweighting algorithm to establish a soil organic matter detection model, and the model is used to realize the detection of organic matter in similar unknown soil samples. The sample of the present invention does not require complex pre-treatment, and has the advantages of fast detection speed, simple and convenient operation, detection stability, and high accuracy.

The invention provides a preparation method of a pump-free SERS (Surface Enhanced Raman Scattering) micro-fluidic chip for detecting mi RNA (Ribonucleic Acid) related to liver cancer and the pump-free SERS micro-fluidic chip. The method comprises the following steps: 1) preparing 20nm gold particles (AuNPs) by adopting a sodium citrate reduction method; 2) respectively modifying Raman signal molecules DTNB and 4-MBA and nucleic acid complementary chains H1 and H2 on AuNPs to prepare AuNPs (at) DTNB (at) H1 and AuNPs (at) 4-MBA (at) H2; (3) modifying the surface of the magnetic core Fe3O4 with nucleic acid aptamers cDNA (complementary deoxyribonucleic acid) 1 and cDNA 2 of the mi R-21 and the mi R-155 to obtain Fe3O4 (at) cDNA 1 and Fe3O4 (at) cDNA 2 (Fe3O4 (at) cDNA); (4) the AuNPs (at) DTNB (at) H1 and the AuNPs (at) 4-MBA (at) H2 prepared in the step (2) are coupled with the Fe3O4 (at) cDNA prepared in the step (3), and Fe3O4 (at) cDNA-AuNPs (at) Raman report (at) H is obtained; and 5) mixing polydimethylsiloxane (PDMS) with a curing agent, adding the mixture into the template to prepare a micro-fluidic chip, and carrying out hydrophilic treatment to construct the pump-free SERS micro-fluidic chip. The method has the advantages of high sensitivity, high specificity, high detection speed and the like.

The invention discloses a high-value target rapid detection and positioning method and system based on an ultra-wide rotary scanning remote sensing image, belongs to the technical field of computer algorithms, and is used for realizing high-value target rapid identification and positioning under the remote sensing image. The method comprises the following steps: acquiring an ultra-wide swing-scan remote sensing image, and timestamps, remote sensing satellite orbit data and swing-scan camera parameters corresponding to each column of pixels of the image; according to the data, calculating an overlapping proportion required by a corresponding pixel position in real time, and segmenting the ultra-wide swing-scanning remote sensing image into a plurality of slices with different sizes; taking a plurality of slices with different sizes as the input of a detector, outputting a target detection frame, and mapping the detection frame back to the ultra-wide remote sensing image according to the positions of the slices in the original image; and taking the pixel coordinate of the center of the target detection frame as the image position of the target, and mapping the pixel coordinate of the target to a WGS84 coordinate to complete the detection and positioning of the high-value target. The method is suitable for rapid identification and positioning of high-value targets under remote sensing images.

The utility model relates to a burr detection device and lamination equipment. The burr detection device comprises a conveying mechanism, a pre-positioning mechanism, a sheet taking mechanism and a visual detection mechanism. The pole pieces are carried by the conveying mechanism and conveyed to the downstream, and are transferred to the pre-positioning platform by the piece taking mechanism before lamination operation. Then, the visual detection mechanism can obtain image information of the edge of the notch of the pole piece so as to judge whether burrs exceed the standard or not. Unqualified pole pieces detected by burrs are removed, and qualified pole pieces are carried to a lamination table by a lamination manipulator to be laminated. The detection cameras on the two sides can simultaneously acquire the image information of the edges of the two notches of the pole piece, and each detection camera can be driven by the driving assembly to move along the edge of the pre-positioning platform, so that the burr detection speed of the single pole piece is relatively high. After the pole piece is sliced, the operations of conveying, burr detection, lamination and the like can be realized in sequence, and all procedures are coherently connected. Therefore, the burr detection device and the lamination equipment can effectively improve the production efficiency.

The invention discloses a method for measuring the content of aluminum in a composite propellant, which comprises the following steps: digesting a sample by using nitric acid and hydrogen peroxide through microwaves, and then fixing the volume and diluting to prepare a sample solution; aluminum element standard solutions with a series of concentrations are prepared, the spectral intensity of the standard solutions is measured through inductively coupled plasma emission spectroscopy, and a standard curve is drawn; and measuring the sample solution under the same condition, calculating the aluminum element concentration through the standard curve, and further calculating the aluminum content in the sample. The method is simple to operate, low in reagent consumption, high in detection efficiency and good in repeatability, has no special requirements on the types of the components of the composite propellant, has universality and can be used for determining the aluminum content in the composite propellant.

The invention belongs to the technical field of wafer automatic detection equipment, and provides a wafer detection point spiral scanning motion device and method in order to solve the problems that linear scanning detection coverage of an existing wafer is uneven, a detection blind area is likely to be generated, and repeated scanning is needed to affect the detection efficiency. The equidistant spiral track movement of the wafer in the plane is completed through the rotating platform and the two transfer platforms, the optical detection module collects spectral information at a fixed position at a preset frequency in the moving process, the whole surface of the wafer is uniformly covered from the center to the edge, no detection blind area is ensured, overlapping or omission in the detection process is effectively reduced, and the detection efficiency is improved. The linear speed is kept at a constant speed, so that re-collection and missing collection of collection points are avoided, point spiral scanning detection in the radial plane of the wafer is completed, the detection device does not need to be started and stopped frequently, the detection speed is adjustable, and the detection efficiency is high.

The invention discloses a small ship detection method based on a convolution block attention mechanism and an explicit visual center. The small ship detection method comprises the following steps: acquiring a water surface ship image; constructing a backbone network model, and inputting the water surface ship image into a backbone network for feature extraction to obtain a semantic information output feature map; a learnable up-sampling operation based on point sampling is carried out on the feature map through a feature pyramid network module, and an up-sampling operation is carried out based on the offset of the feature map to obtain a high-resolution feature map; performing explicit visual center post-processing on the high-resolution feature map to capture a long-distance dependency relationship between pixels and restore edge information to obtain a final feature map; and carrying out splicing and down-sampling on the final feature map again to generate a deep feature map with low resolution and high channel number, and combining the deep feature map with the final feature map to carry out target detection to obtain a central point coordinate of the small ship, a width value and a height value of a detection frame, and category and confidence information of the target.

The invention discloses an online detection method and device for a broken line fault of a lightning conductor of a fan blade and a medium. The device comprises an energy taking module, a pulse generation module, a signal detection module and a remote monitoring system. The method comprises the following steps: 1) the signal detection module receives a pulse signal transmitted by the lightning conductor and records the pulse signal as a pulse response signal; and the signal detection module transmits the pulse response signal and the blade code to a remote monitoring system. (2) the remote monitoring system processes the pulse response signal, if the amplitude of the pulse response signal is smaller than 15 v or the pulse width does not belong to the range [470 ns, 530 ns], it is judged that the lightning conductor of the current blade has a broken line fault, and otherwise, the step (3) is executed; and 3) the remote monitoring system analyzes the pulse response signal and the pulse signal, and if the waveform correlation coefficient of the pulse response signal and the pulse signal is less than 0.9, the lightning conductor is judged to have a broken line fault. According to the invention, the broken line fault condition of the lightning conductor of the fan blade after lightning stroke can be detected on line.