2136 results about "Image extraction" patented technology

The invention relates to the technical field of industrial component detection, in particular to a transformer iron core detection method based on computer vision, which comprises the following steps of: acquiring an iron core image, extracting key pixel characteristics, screening a directional scattering abnormal region to generate an interference map, extracting a consistent gradient region correction image to generate a reconstruction map, and positioning a symmetric disturbance generation structure map by integral gray difference. And analyzing an overlapping relation by a superposition structure graph to generate an abnormal component graph, and evaluating a risk level by matching a reference index to generate an early warning graph layer. Interference reflection and structural features can be distinguished through linkage analysis of the pixel direction vector and the brightness change frequency, correction of a distorted area in an image is realized based on a gray statistical stable value, and the distortion of the image is corrected by constructing a symmetric point map and analyzing the change trend of a gradient difference value sequence. And the structural overlapping relation is quantitatively judged by combining a component mapping profile diagram, so that the relevance between an abnormal region and a key component is clearly expressed, and the grading evaluation capability of various fault risks in the iron core is improved.

The invention discloses a grabbing control system for robot visual guidance, and belongs to the technical field of grabbing control systems.The grabbing control system comprises a visual perception module used for collecting workpiece images, extracting workpiece feature information and calculating accurate coordinates of workpieces in space; the robot main body module is used for receiving a system instruction to drive a carrier to move to a target position; the grabbing execution module is used for being in contact with the workpiece, grabbing the workpiece stably and transferring the workpiece to a designated position to be placed without damage; the system integration module is used for guaranteeing the safety, operability and stability of the feeding and discharging process; according to the system, workpiece information can be accurately collected, the posture of the grabbing execution end is analyzed based on the workpiece position, the grabbing posture is corrected in real time according to the grabbing dynamic state, and the workpiece grabbing precision and stability are improved.

The invention relates to the technical field of image processing, in particular to an image acquisition and analysis method and system, and provides the following scheme: obtaining a visible light and near-infrared multispectral image, generating a spectral difference image, and performing weighted fusion to obtain a first image; segmenting a target region based on the fused saliency map, and calculating a pixel reflectance ratio; solving a color mapping matrix according to the reflectance ratio, and carrying out color correction on the target region to obtain a standardized feature image; and extracting characteristic parameters such as spectrums, colors and textures, inputting the characteristic parameters to a multi-branch convolutional neural network, fusing the characteristic parameters through an attention mechanism, and outputting a state classification result and a quantitative index. The cross-spectral imaging difference can be adaptively compensated, and the fusion precision and the analysis stability are improved.

The invention discloses a PDF (Portable Document Format) document content processing method and device, equipment, a storage medium and a program product, and relates to the technical field of document structured processing. Preprocessing the PDF document to obtain a to-be-processed data set corresponding to each page of the PDF document; determining the page type of each page of the PDF document based on all the to-be-processed data sets and the image of each page of the PDF document; and based on the to-be-processed data set corresponding to each directory page and the image of the directory page, extracting a hierarchical structure relationship of each title data in the directory page, and constructing a directory tree. And matching the title data of the directory page and the title data of the non-directory page based on the semantic similarity and the text similarity between the title data of the directory page and the title data of the non-directory page, and correspondingly filling the content data under each title node of the directory tree according to a matching result to obtain a structured representation result of the PDF document. According to the method and the device, the semantic reduction degree and the structural quality of the PDF document are improved.

The invention relates to the technical field of electronic component detection, in particular to an IC (integrated circuit) carrier plate detection method based on surface state image extraction, which comprises the following steps: acquiring a gray image, analyzing structural parameters, extracting gradient features, detecting boundary disturbance, integrating the image, calculating an abnormal score, identifying a defect position area, extracting features and outputting an identification result. According to the invention, by analyzing the structure parameters of the bonding pad in the gray level image, calculating the edge line segment, the center coordinate and the spacing, and constructing the two-dimensional coordinate system, the regional positioning reference is enabled to have geometric consistency, the coordinate mapping is combined with the gradient direction change frequency and the continuous aggregation point, the boundary disturbance identification precision is improved, and the image division is executed based on the disturbance region. According to the method, non-functional region mixing is effectively avoided, a clustering and probability model is introduced after region gray level statistics, a deviation scoring mechanism is constructed, gray level feature abnormity is accurately recognized, the discrimination capability of small-amplitude and low-contrast defects is improved, and the selectivity and target focusing performance of feature detection are enhanced.

The invention relates to the technical field of unmanned aerial vehicle detection and recognition, in particular to an unmanned aerial vehicle detection system and method based on vision, laser radar and sound waves, and the system comprises a sensor unit, a data processing fusion unit and a target recognition tracking unit. The data processing fusion unit performs multi-source data alignment and weighted fusion to extract texture, shape and color features of the image, point cloud generates a depth map, sound wave signals are mapped into a two-dimensional feature map, and weights are dynamically distributed based on sensor reliability to generate a unified feature map; the target identification and tracking unit identifies the type of the unmanned aerial vehicle by using the optimized deep learning model and realizes accurate prediction and updating of position and speed states in combination with Kalman filtering, and the decision response unit triggers sound-light alarm and wireless alarm for the non-cooperative unmanned aerial vehicle in real time and transmits target dynamic information to the ground station. And the target detection accuracy of the unmanned aerial vehicle in a complex environment is improved.

The invention relates to the technical field of automatic control, in particular to a multi-parameter cooperative control method and device for a clamping system of a precision boring and milling machine. According to the method, an industrial camera is used for collecting and processing a workpiece image and extracting visual features, and the visual features are fused with process parameters of a workpiece and geometric distribution parameters of a workpiece key area obtained from a process database; generating a comprehensive state vector; based on the vector, outputting a target clamping force adjustment coefficient by a pre-trained MLP decision model, and calculating a final target clamping force in combination with a safe clamping force range; in the main clamping stage, vibration, temperature and pressure signals are synchronously collected, multi-modal disturbance characteristics are extracted, observation vectors are constructed, the observation vectors are input into a preset fuzzy rule base for reasoning, and compensation activation factors are obtained; and finally, a current correction instruction is dynamically generated according to the clamping force deviation and the compensation factor, and an electro-hydraulic proportional valve is driven to achieve accurate compensation. According to the invention, the stability and the machining reliability of the clamping system are obviously improved.

The invention discloses a target detection method and device based on an aerial view, equipment, a medium and a product. The method comprises the following steps: extracting multi-scale features for an input image of a multi-view-angle surround-view camera; converting the point cloud data of the laser radar into a sparse voxel structure and generating BEV features for voxel features; fusing the multi-scale features and the BEV features to obtain fused BEV features; road topology and semantic information are extracted according to the fused BEV features, and a map is generated; and determining a three-dimensional bounding box and a motion state of the traffic participant according to the fused BEV features and a prediction result of the position of the traffic participant in the map. According to the technical scheme, multi-scale geometric features, BEV features and semantic information are fused, time sequence features of dynamic target detection and static map segmentation are synergetic, the reliability of target detection based on aerial view is improved, and the method is suitable for multi-task BEV perception of three-dimensional object detection, high-precision map segmentation and motion prediction.

The invention relates to the technical field of multi-source data fusion, in particular to a multi-source geological data processing method and system of a three-dimensional geological model.The method comprises the following steps that mountain landform and river valley images are obtained, gray frequency characteristics are extracted, a frequency energy gradient layer is constructed, a frequency continuous response area is screened to generate a structure boundary set, and a structure boundary set is constructed; the method comprises the following steps of: extracting a boundary normal vector by utilizing principal component analysis, identifying boundary sections with consistent directions, estimating a physical property parameter gradient direction, judging an included angle screening blocking region, generating a space attribute limiting layer, carrying out space alignment analysis on an overlapping region vector included angle, updating a boundary label, and generating an available attribute path structure set in three-dimensional geological modeling through a Dijkstra algorithm. According to the method, a conduction model is constructed through frequency domain decomposition and logarithmic transformation enhanced recognition, frequency window analysis noise reduction, principal component extraction vector analysis direction and center difference estimation, dynamic matching is promoted through alignment, a Dijkstra algorithm optimizes a path, and the geological model bedding characterization and conduction simulation precision is improved through cooperation of a multi-dimensional technology.

The invention discloses an ecological environment detection method and system based on multispectral remote sensing fusion, and relates to remote sensing image processing. The method comprises the following steps: collecting multispectral remote sensing image data of a target area; performing multiband joint atmospheric correction processing on the multispectral remote sensing image according to the scattering coefficient, the atmospheric light value and the transmissivity of each band; performing foreground and background analysis on the corrected multispectral remote sensing image; fusing the vegetation area and the non-vegetation area of each wave band image by adopting different weight strategies to generate a multispectral fusion image; and extracting spectral features of the multispectral fusion image, constructing a standard vegetation spectral feature library, and identifying regions deviating from a standard vegetation spectrum through an anomaly detection algorithm according to the extracted spectral features to obtain various vegetation coverage rates. In view of low vegetation identification precision caused by direct foreground and background division of a multispectral remote sensing image under an atmospheric interference condition, vegetation division is performed after a clear image is obtained, so that the detection precision is improved.

The invention discloses a thermal printing image processing method, device and equipment and a medium, and relates to the field of thermal printing. The method comprises the following steps: acquiring an original image to be printed, and converting the original image into a grayscale image; calculating a gradient difference value of the grayscale image, and generating a threshold distribution map; generating a heat accumulation risk map based on an average gray value in a preset neighborhood window taking each pixel point of the gray image as a center; generating a preliminary binary image according to the threshold distribution map; extracting image features of the preliminary binarization image, and classifying the original image to obtain an image classification result; calling a target processing parameter set corresponding to the image classification result from a plurality of preset processing parameter sets according to the image classification result; performing optimization processing on the preliminary binary image according to the target processing parameter set and the heat accumulation risk map to obtain an optimized binary image; and the final printing image adaptive to the resolution of the target printer is generated, so that the printing definition is improved.

The invention provides a science and technology information deep analysis method and system based on cross-modal semantic enhancement, and relates to the technical field of science and technology information analys.The method comprises the steps that firstly, a cross-modal semantic anchor point set is constructed, and the cross-modal semantic anchor point set comprises text theme anchor points extracted from science and technology information texts, visual object anchor points extracted from images and the association mapping relation of the text theme anchor points and the visual object anchor points; constructing a semantic conduction path between anchor points based on the cross-modal semantic anchor point set, realizing bidirectional information transmission, generating a cross-modal semantic enhanced representation, performing hierarchical semantic analysis on the enhanced representation to obtain a topic association rule, a technical element dependency relationship and a concept evolution sequence, and integrating the topic association rule, the technical element dependency relationship and the concept evolution sequence into an analysis conclusion; the analysis conclusion is reversely mapped to adjust the association mapping relation strength, an updated set is obtained, finally, a structured science and technology information analysis report is generated based on the updated set, logic connection of all modules is achieved, and comprehensive and accurate science and technology information analysis is provided for users.

The invention discloses an industrial robot image processing method based on image fusion, and relates to the technical field of intelligent aquaculture, and the method comprises the steps: collecting an original image in real time through deploying an image collection device integrating visible light, polarization and multispectral imaging, and extracting suspended matter density, water body light transmittance and illumination intensity change information; generating a first image set; suppressing suspension interference through image filtering and enhancement processing to obtain a first corrected image; extracting an aquatic product individual region, performing multi-source image fusion, identifying color deviation, texture interruption and reflection feature anomaly regions, and constructing a lesion candidate set; gray scale reconstruction, edge gradient and brightness normalization correction of a multispectral channel are executed based on illumination and reflection changes, and a high-quality fusion image is generated; and calculating a health anomaly probability coefficient of the target individual by using the depth recognition model, comparing the health anomaly probability coefficient with a threshold value, and recording a recognition result and collecting information if the threshold value is exceeded. The method can significantly improve the accuracy of aquatic individual lesion recognition.

The invention relates to the field of fatigue driving behavior analysis based on image recognition, in particular to a fatigue driving behavior feature extraction and analysis method based on image recognition, which comprises the following steps of: acquiring an initial state set of a driver in real time through an IMU (Inertial Measurement Unit), an RGB (Red, Green and Blue) camera and an MEMS (Micro Electro Mechanical System) vibration sensor, converting the initial state set into four images such as a head attitude angular velocity oscillogram, extracting image features and inputting the image features into corresponding preset models to obtain fused feature data, dynamically adjusting weights through scene context features, calculating cognitive load indexes and dividing processing modes; according to the method, multi-modal data fusion and dynamic weight adjustment are realized, the cognitive load of the driver can be accurately evaluated and graded intervention can be performed, and the driving safety is improved.

The invention belongs to the technical field of defect identification, and particularly relates to a gas turbine blade defect identification method based on an improved YOLOV8 network. Comprising the following steps: capturing a defect image; effective information is extracted from the image or subsequent target detection, classification and segmentation task requirements are met; performing data enhancement through geometric transformation and color perturbation; adding structured labels or annotations to the images, and endowing the images with semantic information; the scheme is improved on the basis of YOLOv8 so as to be realized in gas turbine blade defect detection, and training of the model is completed in a distributed heterogeneous computing framework by using an acquired training data set and an acquired verification data set and is used for a prediction task. The method can achieve the pixel-level detection and positioning of the defects of the blade, can effectively meet the daily detection demands, and gives consideration to the detection precision and real-time performance.

The invention discloses a zero sample anomaly detection method and system based on a triple perception learning enhanced visual language model, and relates to the field of computer vision, and the method comprises the steps: extracting global and local visual features from an input image; in the visual coding process, local features in a deep network are corrected through a spatial perception attention enhancement module, and fine-grained attribute text description is generated for abnormal visual features; performing deep semantic alignment on the attribute text description and the general text prompt through an attribute perception guide module; calculating the similarity between the enhanced visual features and the optimized text features, and generating a pixel-level abnormal segmentation map; in the inference stage, the segmented image is converted into a space attention weight through an anomaly perception reconstruction module, the space attention weight is fed back to a visual encoder to generate final global feature representation, and an anomaly score is calculated. According to the method, under the condition that a target domain training sample is not needed, the anomaly detection and positioning accuracy and generalization ability of the model under the scenes of industrial defect detection and the like are remarkably improved.

The invention discloses a medical image registration method based on large model robust features, and the method comprises the following steps: constructing a large model registration network which comprises a structure perception feature encoder module and a pyramid deformation field prediction module; acquiring a medical input image, and performing feature extraction on the medical input image through a structure perception feature encoder module to obtain image extraction features; performing deformation and registration on image extraction features based on a pyramid deformation field prediction module to obtain a medical registration image; and training based on the complete loss function pair to optimize the registration model. According to the method, the registration network SAMIR is utilized, universal visual features across anatomical regions can be effectively extracted, the registration accuracy is improved, the anatomical rationality of a deformation field is improved, the migration potential of a natural image pre-training model to a medical registration task is proved in the absence of medical priori knowledge, a feature-level loss function is proposed, and the medical registration efficiency is improved. And the registration consistency is further enhanced.

The invention relates to the technical field of path control, in particular to a combined path and laser processing planning method based on machine vision, which comprises the following steps: acquiring an image by an industrial camera to extract contour measurement parameters to identify defective materials, setting path start-stop optimization connection smooth obstacle avoidance based on a feature table, and detecting a dynamic adjustment range of a hot area. And according to the adaptive group setting threshold evaluation capability, the delay prediction adjustment power speed is calibrated, an instruction set is called to compensate, predict and correct error dynamic matching, and an optimal processing path and a synchronous processing setting table are output. According to the method, the workpiece image is obtained in real time, the machining features are analyzed, the machining precision and efficiency are improved, fine control in the machining process is achieved, effective cooperation of path planning and machining parameters is ensured, the problem of mismatching of path planning is avoided, the response precision and stability in the machining process are improved, and the machining precision and efficiency are improved. The processing quality and time management are remarkably improved, and low efficiency and quality fluctuation caused by manual adjustment in the prior art are avoided.

The invention relates to the technical field of disease and insect pest image analysis, in particular to an AI image recognition and grading method for field crop leaf disease and insect pests, which comprises the following steps: under the irradiation of a field fixed light source, synchronously acquiring a plurality of polarized reflection images around crop leaves at preset angle intervals; extracting a pixel polarization degree matrix of a leaf area in each polarization reflection image; inputting the pixel polarization degree matrix into a polarization transmission model, outputting a cuticle anomaly coefficient graph, and marking an area exceeding a preset anomaly threshold in the cuticle anomaly coefficient graph as a highlight display area; matching an infection type template library according to a highlight display area distribution mode in the abnormal coefficient graph; and calculating an infection intensity value by combining the diffusion gradient of the highlight area, and outputting a pest grade. According to the method, the boundary of the optical mutation region of the focus region is depicted, so that the physical interpretation of disease detection is improved, and distinguishable feature spaces are provided for different infection mechanisms (such as fungal growth layers and insect pest piercing and sucking points).

The invention provides a community open space psychological recovery effect assessment method and system based on multi-modal perception. The method comprises the following steps: acquiring a physiological signal, a movement track, eyeball movement data and a panoramic image of a user; analyzing the image, extracting environmental element space distribution parameters, and calculating an environmental information entropy value; according to the physiological signal, calculating physiological stress deviation as a first type of error, analyzing the matching degree of a moving track and an environment structure as a second type of error, and combining eye movement characteristics and visual attraction distribution to calculate attention deviation as a third type of error; inputting the three types of errors and entropy values into an evaluation model, and outputting a recovery efficiency index and a multi-dimensional index representing error mitigation; and associating the environment parameters with the multi-dimensional indexes, explaining and extracting an environment intervention critical value and a recovery effect function, and generating a space optimization evaluation report in combination with dynamic indexes. According to the method, quantitative evaluation and accurate optimization decision support of the psychological recovery effect of the open space of the community are realized.

The invention discloses a glacier area surface water resource distribution image extraction method based on deep learning, and belongs to the field of surface water resource image extraction, and the method comprises the steps: obtaining a multispectral remote sensing image covering a glacier area and digital elevation model data, and carrying out the data preprocessing; building a YOLOv12 instance segmentation model comprising a backbone network, a neck network and a detection head, and constructing an improved YOLOv12 segmentation model; carrying out model training and optimization; segmenting and extracting surface water resources through the optimized segmentation model; processing is performed after segmentation is completed, and finally an optimized multi-class surface water resource distribution diagram is generated. According to the method, based on an improved YOLOv12 network architecture, the glacier area surface water resources in the remote sensing image are comprehensively and automatically extracted by fusing vision, a convolution block attention module and a space-to-depth convolution technology, the problems of high reflection of ice and snow, ice cracks, complex stone glacier textures and fuzzy edges are effectively solved, and the segmentation precision is improved.

The invention discloses a safety real-time detection method in a complex scene based on multi-scale feature fusion, and relates to the technical field of safety detection, and the method comprises the steps: S1, obtaining a to-be-detected complex scene image; s2, extracting multi-scale initial feature maps with different semantic information and spatial details; s3, inputting the initial feature maps of the plurality of scales into an adaptive feature fusion network; s4, inputting the enhanced feature pyramid into a lightweight decoupling detection head, and executing target classification and bounding box regression in parallel; and S5, based on the category and position information, generating and outputting a final security detection result. The method has the advantages that semantic and detail features of different levels are effectively integrated through a self-adaptive gating fusion mechanism and multi-scale context aggregation, and the detection precision and robustness of the model on a multi-scale target in a complex scene are remarkably improved.

One variation of a method includes: accessing an image of an inventory structure captured by a robotic system while navigating through a store; detecting a slot in the inventory structure in the image; based on features extracted from the image, identifying a set of product units of a first product type occupying the slot; accessing a target product type assigned to the slot by a store representation; based on features extracted from the image, identifying an electronic shelf label depicted in the image, corresponding to the slot, and advertising the target product type; and, in response to the first product type differing from the target product type assigned to the slot, accessing a set of product data corresponding to the first product type from a product database, and, transmitting the set of product data to the electronic shelf label for rendering within an electronic display of the electronic shelf label.

The embodiment of the invention provides an optical measurement method and system based on camera calibration and a distortion model, and a medium. The method comprises the following steps: shooting a plurality of calibration images and distortion measurement images at different angles and positions based on a three-dimensional calibration workpiece; analyzing the calibration image and the distortion measurement image based on a calibration algorithm to obtain camera parameters and a distortion model; obtaining a to-be-measured object image based on the calibrated camera, and extracting features of the to-be-measured object image; performing distortion correction on the image features of the to-be-measured object based on a distortion model to obtain corrected image features of the to-be-measured object; performing three-dimensional reconstruction on the corrected image features of the to-be-measured object to obtain a measurement image, and analyzing parameter information of the to-be-measured object based on the measurement image to obtain a measurement result; parameter calibration and distortion correction are carried out on the camera, so that the distortion error is reduced, and the measurement precision is improved.

The invention provides a space measurement and analysis method and system for a neutron diffraction rocking curve of a single crystal material, and the method comprises the steps: S1, building a laboratory coordinate system and a sample coordinate system, and setting a test point grid; s2, collecting a two-dimensional diffraction image of each test point to obtain angle information of the test point; s3, integrating the angle information of the test points based on the test point grid, and preprocessing the two-dimensional diffraction image; s4, extracting key structure parameters of each test point based on the preprocessed two-dimensional diffraction image; s5, constructing a stress field according to the key structure parameters; and S6, constructing a distribution map of crystal internal lattice parameters in a three-dimensional space according to the stress field. The method provided by the invention effectively improves the capability of identifying the non-uniformity and local distortion of the crystal structure, can be widely applied to the fields of residual stress evaluation, defect evolution monitoring, material quality control and the like, and has a remarkable engineering application value.

The invention relates to an improved YOLOv8 small target detection method based on SENet, and the method comprises the steps: introducing semantic dilution loss, and measuring the dilution degree of small target features in a channel; a suppression reverse weight is generated through an SENet structure, and background redundancy is suppressed while a high response area is reserved; a C2f structure of YOLOv8 is fused, and decoupling characteristics are transmitted in cross-layer connection; in the Neck feature pyramid, a SENet response migration relation is constructed; calculating a weight deviation value, and judging whether the small target response has spatial deviation or not; position balance loss is introduced, and feature repositioning is carried out on a small target area with overlarge center-of-gravity drift; designing a channel response consistency measurement index, and measuring the SE response consistency of the small target between different epochs; sENet channel output is extracted from the image, and the variance of target area channel response distribution is counted; a channel with high variability is weakened or suppressed from a current image detection path, a detection head is introduced into a channel gating mechanism, a stable channel is adaptively selected to participate in prediction, and the small target feature representation capability is significantly enhanced.

According to the mass center self-balancing chassis system for U-rib weld defect detection and the control method, the problem that when detection equipment moves on a U-rib curved surface, the precision is reduced due to mass center deviation and view field inclination can be solved. The system integrates a Mecanum wheel omni-directional moving unit, four groups of electric telescopic supporting legs, a two-axis holder and an ECU (electronic control unit), and has the innovation point that dynamic mass center balance and posture self-correction are realized through rigid-flexible combined supporting leg design and multi-modal sensing fusion. The ECU adopts an improved Canny algorithm to process the image, extracts a U rib edge point set and fits a reference center; in combination with pressure and tilt angle sensor data, a real-time mass center is solved through a moment balance formula, and then an improved self-adaptive PID algorithm is used for driving the supporting legs to stretch out and draw back, so that the mass center returns to a reference center; the two-axis holder compensates the view field angle according to the inclination angle data to ensure that the view field angle is perpendicular to the weld joint. Through cooperation of a mechanical structure and an algorithm, the system can control centroid offset to be smaller than or equal to 5 mm and view field inclination to be smaller than or equal to 1 degree, the detection stability and precision are remarkably improved, and the algorithm is easy to deploy.

The invention discloses a visual filter tip modeling recognition and control system and a data interaction method thereof, and relates to the technical field of image recognition and intelligent control, and the method comprises the following steps: carrying out the continuous image collection of a filter tip through an industrial camera, and obtaining an image sequence of the filter tip in a conveying process; for each frame of image in the image sequence, extracting a transverse brightness change curve in a gray domain, and constructing an image brightness distribution matrix based on the spatial distribution of the filter tip in the image recognition area; according to the invention, by introducing a frequency domain analysis and risk prediction control mechanism, intelligent perception, quantitative evaluation and dynamic suppression of periodic optical interference are realized. The system can extract interference frequency characteristics, calculate the stability and occupancy ratio, predict and identify the risk in combination with a historical misjudgment model, dynamically adjust the duty ratio of the LED light source according to the risk result, break the resonance relation with the conveying rhythm, and suppress false characteristic interference. The method improves the accuracy, stability and yield of an identification system, and has good engineering application value.

The invention discloses a method for estimating plant biomass based on map multi-modal feature extraction and fusion. The method comprises the following steps: acquiring a plant RGB image and a multi-spectral image; inputting the preprocessed RGB image and NIR wave band image into a double-model cooperation segmentation framework to realize image segmentation; binary image features, color features, texture features, reflectivity and the like are calculated, feature splicing is carried out, and high-dimensional features are constructed; carrying out dimension reduction on the high-dimensional features; and training a deep neural network through the effective features and the biomass to realize biomass estimation. According to the method, a zero sample learning-based double-model cooperation segmentation framework is utilized to realize accurate segmentation of a single plant on the premise that a large number of training sets are not needed; multi-modal feature information is extracted based on the segmented single plant image, an improved SHAP model is introduced to reduce the feature space dimension, and the inversion precision and the operation efficiency are improved while the information effectiveness is ensured; through a high-precision deep neural network model, rapid, lossless and accurate biomass acquisition is realized.

The invention provides an image region-of-interest extraction method and system based on a Mama architecture, and relates to the technical field of image processing, and the method comprises the steps: obtaining a to-be-extracted image; performing multi-scale feature extraction on the to-be-extracted image through a local enhancement module; multi-scale semantic enhancement features are generated through a cross-scale self-attention module, and dimension reduction processing is performed through a feature conversion module; local detail features are generated through an adaptive detail enhancement module; performing global context enhancement processing on the local detail features through a pyramid pooling module to generate context enhancement features; performing up-sampling processing on the context enhancement features; the context enhancement features after up-sampling processing are fused through a self-adaptive global-local fusion gating module; and carrying out image extraction based on the decoded fusion features. The image segmentation precision is improved, the model is light in weight, reasoning is fast, and the problems of image boundary blurring and scale variability are effectively solved.