646 results about "Hyperspectral imaging" patented technology

The invention discloses a carbon fiber composite material surface modification spraying system and a control method thereof. The system comprises a multi-axis robot, a plasma spray gun, a contact angle measuring probe, a 3D line laser scanner, a hyperspectral imager, an environment sensor and a controller. According to the method, the plasma power and the robot speed are adjusted in real time through contact angle measurement and plasma treatment feedback control, so that the CFRP surface can accurately reach a target value, and the coating adhesive force is improved; 3D line laser scanning and hyperspectral imaging are combined, and the posture, the distance, the wet film thickness and the component uniformity of the spray gun are monitored in real time; based on a prediction model fusing a physical model and a neural network, a self-adaptive fuzzy PID control algorithm is adopted, and the coating flow and the track posture of a spray gun are accurately regulated and controlled in real time. The problems of weak coating binding force, uneven thickness, orange peel, sagging and serious coating waste in traditional spraying are effectively solved, and self-adaptive, high-quality and green spraying of workpieces with complex curved surfaces is achieved.

The invention discloses a marine plankton hyperspectral imaging detection system, relates to the technical field of spectrum detection, and is used for solving the problem of poor spectrum classification and recognition in a water body environment. According to the method, the hyperspectral image and the environmental disturbance parameters are synchronously acquired, and the disturbance mapping sequence is constructed, so that accurate alignment between the image and disturbance is realized. A disturbance contribution weight matrix is constructed based on a dominant wave band, pixel-level spectrum stripping is carried out, a plankton purification spectrum is extracted, the spectrum purity and the recognition accuracy are improved, and a spectrum abnormal drift region is precisely recognized in combination with derivative spectrum change and image texture features; a coupling relation between the drift region and background disturbance is further established, a dynamic interference factor map is generated, and identification model parameters are dynamically adjusted, so that the identification stability and classification accuracy of the system in a complex interference environment are improved, and the method is suitable for multi-scene marine ecological monitoring.

The invention relates to the technical field of rice grain detection, in particular to a rice grain mixed-planting rapid nondestructive detection method based on hyperspectral imaging, which comprises the following steps: S1, collecting a rice hundred-grain board hyperspectral image, carrying out preprocessing and region-of-interest segmentation, and extracting a full-wave band spectrum grayscale image of a target region; s2, constructing a multi-channel attention mechanism network model suitable for nondestructive testing of mixed rice grains, and inputting the full-wave band spectrum grayscale image as a data set for classification training; according to the rapid nondestructive testing method for mixed planting of the rice grains based on hyperspectral imaging, spectrum and spatial characteristics of the rice grains are fused, spectrum and spatial characteristic information of the rice grains is enhanced through a multi-channel attention mechanism and a self-attention mechanism module, the classification precision of the rice grains is further helped to be improved, and the detection accuracy of the mixed planting of the rice grains is improved. According to the multi-channel attention mechanism network model provided by the invention, a segmentation module and a classification module are fused, so that the parameter quantity during high-resolution image training can be reduced.

The invention discloses a rapid nondestructive testing method and system for the lipid content and deterioration degree of red pine nuts based on hyperspectral imaging and deep learning, and belongs to the technical field of nondestructive testing of the quality and safety of agricultural and forestry products. The method is provided for solving the problems that an existing method for detecting the lipid content and the oxidation degree in the red pine nut kernels is generally complex in operation process, high in large-batch detection cost, long in consumed time and difficult to achieve detection in the whole storage and transportation process. The method is characterized by comprising the following steps: acquiring original near infrared spectrum data of a pine nut sample through a collected hyperspectrum; determining the lipid reference content truth value and the oxidation deterioration reference degree of the pine nut samples at different sampling times, and establishing a database according to the values; the method comprises the following steps: preprocessing collected original near infrared spectrum data of red pine nuts, and dividing red pine nut sample data collected in different batches into a training set and a verification set; and designing an improved one-dimensional cavity convolutional network based on a dynamic weight distribution module to construct a deep learning model, wherein the deep learning model is used for constructing a deep learning structure suitable for spectral feature analysis of the red pine nuts. And a back propagation algorithm is adopted to train the constructed model, and reverse updating of network weight parameters is realized by minimizing a loss function. When the performance of the constructed model meets the rapid detection requirement, the method is used for efficient and lossless synchronous detection of the lipid content and the oxidative rancidity degree of the red pine nuts.

The invention relates to the field of optical nondestructive testing, and discloses a visualization method for detecting mango quality based on hyperspectrum, which comprises the following steps of: firstly, selecting mango samples, and simulating different mango ripening stages by dividing the mango samples into a plurality of groups of samples; acquiring a high-dimensional spectral image by using a hyperspectral imaging system, and measuring the hardness, the soluble solid content and the titratable acid content; correcting and extracting the high-dimensional spectral image to obtain and optimize the average reflection spectrum of the sample, and establishing a mango maturity classification model and a quality regression prediction model; performing characteristic wavelength contribution degree analysis and screening, and optimizing a quality regression prediction model by adopting a key wave band; the method is applied to each pixel point of a high-dimensional spectral image, a two-dimensional distribution diagram is generated, and mango quality visualization is achieved. The method solves the problems that in the prior art, rapid and lossless judgment of the mango ripening stage cannot be achieved, accurate prediction and visual distribution of key internal quality indexes cannot be achieved, and intelligent grading and quality control of picked fruits are not ideal.

The invention discloses a hyperspectral image super-resolution method and device based on a physical diffusion model. The method comprises the following steps: acquiring a low-resolution hyperspectral image and a corresponding high-resolution multispectral image; extracting edge information and semantic information from the high-resolution multispectral image; and inputting the low-resolution hyperspectral image, the edge information and the semantic information into a physical diffusion model, and reconstructing to obtain a high-resolution hyperspectral image. According to the invention, through a physical information guiding sub-network design, a physical information extraction and fusion method and a physical model guiding sub-network design, a physical mechanism of hyperspectral imaging and a scene physical attribute are successfully and deeply fused into a core iteration process of a diffusion model; the fusion significantly improves the spatial detail definition, spectral fidelity and overall physical consistency of the high-resolution hyperspectral image, and provides more effective physical guidance information and a more robust physical constraint optimization mechanism.

The invention discloses a multispectral and hyperspectral imaging proportion balance optimization method based on satellite remote sensing, and relates to the technical field of balance optimization, and the method comprises the following steps: collecting a first image data set which comprises a low-resolution multispectral image and a hyperspectral image; preprocessing the first image data set, screening out a low-resolution multispectral image from the preprocessed images, improving the spatial resolution of the multispectral image by adopting a super-resolution reconstruction algorithm, and generating a reconstructed multispectral image; fusing the reconstructed multispectral image and the preprocessed hyperspectral image, and performing optimization processing on a fusion result to generate an image with an optimal resolution; and extracting second image data based on the optimal resolution image, and establishing a proportion balance optimization model. According to the method, the data processing time and the imaging cost are accurately calculated, and the image fusion quality evaluation is combined, so that the optimal allocation of the imaging ratio under limited resources is realized.

The invention relates to an internet-of-things monitoring and early warning system for lake water quality. The internet-of-things monitoring and early warning system comprises a multi-mode intelligent sensor acquisition system, a self-adaptive wireless transmission system, a distributed data storage system, an intelligent data analysis system, a visual interaction system, an intelligent early warning system and a remote intelligent control and repair system. The invention also relates to a monitoring method using the lake water quality Internet of Things monitoring and early warning system. Traditional and innovative sensors are cooperated to comprehensively collect water quality and ecological data, the biosensor accurately detects specific pollutants, the hyperspectral imaging sensor monitors algae distribution and water color change, rich and high-precision data are provided for lake water quality monitoring, the monitoring comprehensiveness and precision are improved, and the water quality monitoring system is suitable for being popularized and applied. The lake ecological protection work is powerfully supported.

The invention provides a compact hyperspectral imager optical system and an imaging method thereof, which are used for solving the technical problems that the overall size of a front telescopic assembly in a traditional hyperspectral imager optical system is difficult to compress, the installation and adjustment difficulty is relatively large, or the incident energy is lost and the space coverage capability is insufficient. And the technical problems of low energy utilization rate and difficulty in improving the sensitivity and the signal-to-noise ratio of the hyperspectral imager of the traditional Dyson type dispersion light splitting assembly are solved. The invention provides a compact hyperspectral imager optical system, which comprises an off-axis two-image-reflection telecentric optical path system based on a free-form surface and a Dyson light splitting system based on a curved surface prism, and realizes high regulation and control capability on light through multiple degrees of freedom of a curved surface shape. Compared with the prior art, the high-spectral imaging device has the advantages that the high-spectral imaging device is simple in structure, high image quality, small distortion and good telecentricity are achieved, meanwhile, the curved prism serves as a dispersion light splitting element, the whole spectrum has high light splitting efficiency, and the signal-to-noise ratio and sensitivity of the high-spectral imaging device can be improved.

The invention discloses a collaborative light regulation and control decision method and system, and relates to the technical field of intelligent regulation and control of plant factory light environments, and the system comprises a plant physiology monitoring unit which is provided with a non-invasive microelectrode array and a hyperspectral imager; and the environment sensing unit is used for collecting illumination intensity, spectrum, temperature, humidity and carbon dioxide concentration data. According to the collaborative light regulation and control decision-making method and system, fusion of multi-source heterogeneous data is realized through a pulse time sequence coding mechanism, and the problem that environmental parameters and plant physiological statuses are difficult to collaboratively analyze in the prior art is solved; based on a dynamic optimization architecture of double-alliance Nash equilibrium, autonomous collaborative decision-making of spectrum parameters can be completed in a short time, and compared with a traditional preset light formula, the response speed is improved; decision effectiveness can be synchronously guaranteed on the virtual model and biological entity level, the light suppression misjudgment rate is reduced, and meanwhile the plant light energy utilization rate is stably improved.

Aiming at the problems of low efficiency, strong subjectivity, insufficient modeling ability, complex hyperspectral data noise, weak waxiness spectrum difference and the like of a traditional rice grain waxiness discrimination method, the invention provides a hyperspectral rice grain waxiness discrimination method based on multi-branch collaborative modeling. The method comprises the following steps: S1, acquiring glutinous and non-glutinous rice grain images by using a hyperspectral imaging system to obtain spectral data; s2, preprocessing spectral data through a combined method of SG smoothing, an asymmetric weighted penalty least square method and multivariate scatter correction to reduce noise and interference; s3, constructing a multi-branch modeling architecture which comprises a CNN local feature extraction module, an SRU spectrum sequence modeling module and a HorNet global high-order modeling module, and outputting a discrimination result through a classifier after multi-branch features are subjected to fusion and pooling attention weighting; and S4, evaluating the performance. According to the method, high-precision lossless discrimination of the waxiness is realized, and a technical support is provided for germplasm screening and quality evaluation in rice breeding.

The invention relates to the technical field of agricultural intellectualization, and discloses a fruit tree growth monitoring system based on multi-sensor fusion, which comprises a multi-mode sensor array, an integrated soil humidity sensor, a three-dimensional ultrasonic ranging sensor, a hyperspectral imaging sensor and a microenvironment meteorological station, and realizes multi-parameter synchronous acquisition; a nonlinear data calibration model adopts a layered architecture, and cross interference between sensors is eliminated; a dynamic weight distribution algorithm is built in the edge computing node, multi-source heterogeneous data are fused in real time, a fruit tree growth situation map is generated, and a low-power-consumption processor is adopted to support lightweight AI model deployment; the self-adaptive power supply module adopts a photovoltaic-temperature difference composite power generation structure to realize cable-free deployment; the wireless ad hoc network communication unit constructs an inter-tree relay network based on a LoRa protocol, supports low-power-consumption remote data return, and integrates a relay node health degree evaluation mechanism. The real-time performance of fruit tree growth monitoring, the data reliability and the situation analysis accuracy are improved.

The invention discloses a coarse cereal aflatoxin detection method combining hyperspectral imaging and deep learning, and belongs to the field of agricultural product quality safety detection. The method comprises the following steps: firstly, acquiring hyperspectral imaging data of coarse cereal grains, extracting space and spectral information, forming basic data, and performing preprocessing, space-time registration and feature alignment; building a hyperspectral space-spectrum double-branch feature fusion convolutional neural network deep learning model, and importing the processed data to complete iterative training; and finally, performing same-standard data acquisition and processing on to-be-detected coarse cereals, importing the trained model, and realizing aflatoxin detection and pollution area positioning through pixel-level analysis. According to the method, deep fusion of space and spectral features is realized, the detection accuracy and generalization ability are effectively improved, the detection result is stable, reliable and traceable, and the fine detection requirement of aflatoxin in coarse cereals is met.

The invention discloses a radiation correction method and system for an airborne hyperspectral imager. The radiation correction method and system are used for improving the radiation correction precision in complex weather. The method comprises the following steps: calculating a theoretical sun position based on unmanned aerial vehicle initial coordinates, ephemeris data and atmospheric parameters; performing course and differential compensation on a solar azimuth angle by combining IMU course and wind speed and direction data, performing refraction correction on an elevation angle, and outputting a high-precision sun position parameter; hyperspectral images and irradiance data are synchronously acquired, quaternion interpolation is combined with IMU data to realize space-time alignment, and attitude disturbance is eliminated; a hemispherical space irradiance model is constructed based on alignment data, cloud layer modes (clear sky, thin cloud and thick cloud) are divided according to atmospheric optical thickness variance, and Lambert body proportion correction, a discrete longitudinal standard method or a deep convolutional spectral network are respectively adopted for radiation correction. The method remarkably improves the correction precision and generalization ability, supports large-area image splicing, and is suitable for quantitative application of geological exploration, environment monitoring and the like.

The invention discloses a laser speckle testing device based on hyperspectral imaging, which realizes high-precision measurement of laser spots and speckle signals under the condition of not depending on an expensive refrigerating system by optimizing an optical structure and a control strategy. The device comprises an imaging lens, a lens electric control module, a liquid crystal tunable optical filter module, an ND filter wheel, a driving mechanism of the ND filter wheel, a two-dimensional image sensor, a main control unit, a three-axis acceleration sensor, a PC host and a tested screen. A noise calibration piece preset in an ND filter wheel is used for collecting noise data in real time before each time of measurement, and noise deduction is carried out before the filter is switched, so that the timeliness and accuracy of data correction are improved; meanwhile, the integrated three-axis acceleration sensor is used for monitoring vibration, and when vibration exceeding a preset threshold value is detected, it can be judged that currently collected data is invalid, and re-collection is automatically triggered. The device is compact in structure, low in cost and suitable for the fields of laser display, projection calibration, high-brightness light source detection and the like.

Various embodiments of the teachings herein include an apparatus for hyperspectral imaging of electrode webs during electrode manufacturing for battery cells. An example includes: a hyperspectral spectroscopy unit with a line scan camera for imaging, the spectroscopy unit configured: to capture hyperspectral images of the forward-moving electrode web at a predefined location of the electrode manufacturing, to ascertain inhomogeneities of the electrode web from the images, wherein an inhomogeneity is a deviation of the chemical composition of the layers or of the particle size distribution from predefined target variables, and to ascertain and save a local position of the inhomogeneities in the longitudinal direction of the electrode web; and a deflection roll over which the electrode web is guided. The line scan camera captures the images at the position of the contact surface between the deflection roll and the electrode web.

The invention discloses a field-of-view self-stabilization and self-adjustment hyperspectral camera system and a control method, and relates to the technical field of hyperspectral imaging. The system comprises a first imaging lens group, a movable slit, a two-dimensional linear driver, a slit outer layer light shield, a collimating lens group, a light splitting element, a second imaging lens group, an area array image sensor, an attitude sensor and a control acquisition circuit. Angle disturbance is obtained in real time, two-dimensional compensation displacement of the slit is calculated to drive the slit to move reversely, field-of-view collection is kept basically constant, collection is triggered according to a set rhythm, slit position and attitude data are recorded synchronously, and the pre-calibrated slit position and mapping of original coordinates and ideal coordinates are combined, so that field-of-view collection is completed. Geometric transformation and resampling dynamic correction are carried out on view field inclination aberration and spectrum bending aberration generated along with slit position changes, the environmental adaptability and reliability are improved, a hyperspectral data cube with accurate space coordinates and a stable spectrum curve is output, and the method is suitable for unmanned aerial vehicle-mounted hyperspectral remote sensing.

The embodiment of the invention discloses a self-propelled hyperspectral imaging control method and system. The method comprises the following steps: based on an ultra-wideband wireless multilateral positioning method, acquiring ultra-wideband positioning information corresponding to a self-propelled hyperspectral imaging platform, and acquiring relative positioning information and a platform high-frequency vibration signal; based on a Bayesian fusion method, carrying out weighted fusion processing on the ultra-wideband positioning information and the relative positioning information; obtaining deformation data corresponding to the platform based on strain gauges arranged at a plurality of key positions of the platform; and according to the fused positioning information, the deformation data and the high-frequency vibration signal of the platform, performing adaptive closed-loop control on the platform so as to complete path correction of the platform, structural deformation offset, intelligent damping design between a vibration sensor and a pneumatic tire, and vibration compensation processing on a hyperspectral imaging module of the platform. According to the embodiment of the invention, the imaging precision is improved while the navigation positioning precision is improved.

An endoscopic imaging system for use in a light deficient environment includes an imaging device having a tube, one or more image sensors, and a lens assembly including at least one optical elements that corresponds to the one or more image sensors. The endoscopic system includes a display for a user to visualize a scene and an image signal processing controller. The endoscopic system includes a light engine having an illumination source generating one or more pulses of electromagnetic radiation and a lumen transmitting one or more pulses of electromagnetic radiation to a distal tip of an endoscope.

The invention relates to the technical field of constructional engineering quality monitoring, and discloses a constructional engineering quality detection system integrating machine vision and artificial intelligence, comprising a multispectral imaging array which comprises an infrared imaging unit and a hyperspectral imaging unit and is installed on an engineering field support, the system is used for collecting surface texture and internal structure spectral data of a building material. According to the constructional engineering quality detection system fusing machine vision and artificial intelligence, through deep fusion of machine vision and artificial intelligence, the technical problems existing in the field of constructional engineering quality detection are solved; the interference layer is effectively penetrated to obtain the intrinsic characteristics of the material; the causal knowledge graph traceability unit accurately locates a defect responsibility chain, the defect evolution confrontation generation unit verifies the authenticity of hidden dangers from the physical law level, and the hidden defect identification accuracy is effectively improved through double verification. Detection, disposal and verification full-process closed-loop control is realized, and the problems of construction period delay and cost out-of-control caused by disjunction of detection and disposal are solved.

The invention relates to the technical field of nutritious food configuration, and discloses a system and a method for configuring personalized nutritious food, and the method for configuring personalized nutritious food comprises the following steps: collecting spectral characteristic data and chemical component data of a food material sample by adopting a hyperspectral imaging and mass spectrum combined technology, and obtaining holographic chemical fingerprint information; processing the collected spectral data by using a multi-scale feature extraction algorithm, extracting multi-level feature information of the food materials, and carrying out standardization and dimension reduction processing on the feature data; according to the method, high-precision prediction of the nutrient content of the food material is realized through combination of hyperspectral imaging and mass spectrometry technologies and a deep learning model, so that the nutrient content prediction accuracy and the configuration scheme accuracy are improved, the nutrient balance achievement rate is also improved, and meanwhile, the limitation of a static database is broken through; real-time nutritional ingredient analysis of any food material is achieved, the method can adapt to individual differences and freshness changes of the food materials, and accurate configuration based on actual food material characteristics is provided.

The invention discloses a shelf life kiwi fruit comprehensive quality quantitative evaluation method based on hyperspectral imaging and a convolutional neural network, and the method comprises the steps: 1, obtaining spectral images of kiwi fruits of different shelf lives through a hyperspectral imaging system, and extracting the original spectral data of a region-of-interest sample; step 2, measuring the soluble solid content, hardness and color parameters of the sample, normalizing the parameters with correlation, extracting a dominant factor combination with an accumulative contribution rate greater than or equal to 80% by using a factor analysis method, and constructing a KCQI (Kiwi berry Comprehensive Quality Index) based on factor load weight; step 3, preferably selecting a characteristic wave band related to the KCQI, and performing sample set division; 4, taking the spectral data of the characteristic wave band as input, and adopting a one-dimensional convolutional neural network to construct a quantitative prediction model of KCQI; and step 5, applying the model to a spectral image of a sample, generating a KCQI spatial distribution visual image pixel by pixel, and visually displaying the comprehensive quality change.

The invention discloses a fire coal calorific value detection method based on a hyperspectral image technology, and relates to the field of coal calorific value detection. The problems that an existing coal quality evaluation method is complex in calculation process and low in detection efficiency are solved. The method comprises the following steps: taking coal samples with known coal calorific value parameters as correction set coal samples; a hyperspectral imaging system is used for collecting calibration set coal samples and corresponding black and white reference hyperspectral image data, and spectral data preprocessing is carried out; constructing a partial least square quantitative analysis model by taking hyperspectral data of the preprocessed calibration set as an input variable and coal calorific value parameters as dependent variables; the method comprises the following steps: acquiring a coal sample to be detected and corresponding black and white reference hyperspectral image data by using a hyperspectral imager, and preprocessing the spectral data; and detecting the calorific value of the to-be-detected coal sample by using the constructed partial least square quantitative analysis model by taking the hyperspectral data of the preprocessed to-be-detected coal sample as an input variable. The method is used for coal calorific value detection.

The invention belongs to the field of power equipment monitoring, and relates to a power equipment overheating positioning method and device based on infrared hyperspectral imaging, and the method comprises the steps: synchronously collecting infrared and hyperspectral multi-modal data, and carrying out the preprocessing of the multi-modal data; based on the preprocessed multi-modal data, complementary features are extracted from the infrared and hyperspectral data, and feature-level fusion is realized through deep learning; based on surface data inversion equipment internal heat source distribution, three-dimensional space positioning is realized; performing overheating type intelligent diagnosis, and distinguishing normal heating and fault heating of the equipment; generating a visual report and triggering graded early warning; and through online learning and model updating, the calculation efficiency and robustness are optimized. The sensitivity and the positioning precision of overheating defect detection are improved, and particularly, the recognition capability of early weak faults is enhanced; the space coordinates and intensity distribution of the internal abnormal heat source can be quantitatively calculated, visual fault position guidance can be provided, and the troubleshooting time is greatly shortened.

The invention provides a polarization hyperspectral water quality monitoring system and method, and the system comprises a tower footing polarization hyperspectral imaging module which is disposed on a building beside a to-be-detected water area, and carries out the area array staring scanning of the to-be-detected water area through a polarization hyperspectral imager, so as to obtain a water body polarization hyperspectral image signal of the to-be-detected water area; the image spectral data processing module is used for deducting dark current and spurious signals from the water body polarization hyperspectral image signals, and calculating the water body spectral reflectivity of the water area to be measured according to the water body polarization hyperspectral image signals by combining a gray value and reflectivity relation model constructed by using the hyperspectral image signals of the gray scale calibration plate; and the water quality parameter spatial distribution inversion module is used for performing inversion on the to-be-detected water quality parameters of the to-be-detected water area according to a pre-constructed water quality parameter inversion model and the water body spectral reflectivity. According to the technical scheme, the integration level is high, the automation degree is high, rapid deployment can be achieved, and all-day continuous unattended monitoring can be achieved.

The invention discloses a two-stage Raman hyperspectral imaging method, which relates to the technical field of spectral signal analysis and comprises the following steps: S0, preprocessing Raman hyperspectral data to remove abnormal values; the method comprises the following steps: S1, taking a wave number where a target characteristic peak is located as a target wave number, and based on a neighborhood of the target wave number in Raman hyperspectral data, performing spectrum denoising on the Raman hyperspectral data by using an adaptive low-rank matrix approximation algorithm to obtain the Raman hyperspectral data after spectrum denoising; and S2, using an improved BM3D method based on a rotating block to carry out image denoising on a Raman image after Raman imaging is carried out on a target wave number in the Raman hyperspectral data after spectrum denoising. According to the method, the neighborhood of the target wave number is used as the target area, the adaptive low-rank matrix approximation algorithm is used for spectrum denoising, then the improved BM3D method based on the rotating block is used for image denoising, Raman hyperspectral fast and efficient imaging is achieved, and the method has the advantages of being good in denoising effect, high in speed and light in weight.

The invention discloses an intelligent dynamic regulation and control method for compaction quality of a high-liquid-limit fly ash roadbed, which comprises the following steps of: establishing a mapping relation among high-spectral characteristics of high-liquid-limit fly ash, a specific surface area index, a total hydroxyl content and a water content through a first neural network model, and establishing an optimal water content prediction model through a second neural network model; the third to fifth neural network models construct the correlation of the compactness, the dielectric constant, the rigidity and the water content; two-stage water content regulation and control of a storage yard and a paving layer are realized through hyperspectral imaging of an optical fiber probe and an unmanned aerial vehicle, and pre-humidification and secondary precise regulation and control are performed in combination with an intelligent spraying system; in the compaction stage, data are collected in real time through a multi-frequency ground penetrating radar and a vibration sensor, compaction degree distribution is inverted, layered evaluation is conducted, and rolling parameters are dynamically adjusted. According to the invention, a hyperspectral technology, multi-source sensing and deep learning are fused, intelligent management of compaction quality is realized, and construction uniformity and quality stability of the fly ash roadbed are significantly improved.

The invention provides a tobacco shred wet mass online detection method based on hyperspectral imaging and time-space fusion, which comprises the following steps: acquiring a hyperspectral data cube of a tobacco shred flow, extracting a tobacco shred area and generating a mask; calculating a standardized moisture index of each pixel point in the mask according to the spectral response values of the moisture absorption wave band and the reference wave band to obtain a standardized moisture index image; based on a set index threshold value, converting the standardized moisture index image into a binary image, and performing morphological cleaning; carrying out wet group identification and marking on the cleaned binarized image to obtain an external rectangular frame of the binarized image; overlapping or adjacent external rectangular frames are combined, the combined external rectangular frames are screened based on the minimum area threshold value, the external rectangular frames which do not meet the area condition are removed, and the final tobacco shred wet mass external rectangular frames are obtained; according to the invention, misjudgment caused by shielding is significantly reduced; the high-speed and real-time online detection requirements of a cigarette production line can be completely met.

The invention relates to the technical field of nondestructive testing, in particular to a corn feed aflatoxin content nondestructive testing method based on deep learning, which comprises the following steps: acquiring a hyperspectral image of a corn feed sample; correcting the hyperspectral image; taking the center of the corrected hyperspectral image as a base point, and extracting a plurality of regions of interest; determining whether the working state of the hyperspectral imaging system is stable based on the confidence coefficient fluctuation value, and adjusting the light source intensity of the hyperspectral image according to the ratio; determining the accuracy of the aflatoxin content predicted value based on the detection accuracy rate, and adjusting the weight attenuation coefficient of the deep learning model according to the difference value; and determining whether the stability of the aflatoxin content detection process is qualified or not based on the detection accuracy fluctuation value, and optimizing the deep learning model hyper-parameters according to the relative difference between the qualification rate fluctuation value and a preset qualification rate fluctuation value. The nondestructive detection accuracy of the aflatoxin content of the corn feed is improved.

The invention relates to the technical field of agricultural intelligent equipment, in particular to intelligent monitoring and weeding equipment for gramineous weeds in a winter wheat field, and aims to solve the technical problem of high missing detection rate and false detection rate of a traditional visual identification method caused by similar forms of winter wheat and gramineous weeds in a seedling stage. The equipment comprises an autonomous mobile platform, a multi-mode sensing assembly, a data processing and decision-making unit and a targeted clearing execution mechanism. The multi-mode sensing assembly integrates a hyperspectral imaging module, a three-dimensional shape measurement module and a polarization characteristic analysis module, and pixel-level data registration is achieved in combination with the cooperative illumination system. The data processing unit performs high-confidence classification decision on the multi-source features through a hierarchical Bayesian fusion model; and the targeted clearing mechanism adopts a laser galvanometer system to carry out non-contact precise clearing on the identified weed growing points. Intelligent identification and efficient physical weeding of weeds in the winter wheat field are achieved, and the method has the advantages of being free of chemical residues and high in operation continuity.