1050 results about "Microscopic image" patented technology

The invention discloses a microscopic image seamless splicing and enhanced reconstruction method based on deep learning, and the method comprises the following steps: S1, collecting a plurality of original images with overlapped regions, and recording the spatial position information and imaging parameters of the original images; s2, preprocessing the original image to generate a standardized image sequence; s3, inputting the standardized image into a structure perception feature extraction network, and extracting a feature map fusing textures and structures; s4, inputting the feature image and the original image into an image registration module; s5, inputting the registration image into the boundary attention splicing network; s6, inputting the seamless image into the residual hierarchy reconstruction network, and enhancing image details through hole convolution and multi-scale branches; s7, image quality evaluation is executed, and the structural similarity, the signal-to-noise ratio and the edge retention rate are calculated; and S8, constructing a training set and carrying out end-to-end training optimization based on a joint loss function. According to the method, a multi-module deep network is fused, and seamless splicing and high-quality enhanced reconstruction of microscopic images are realized.

The invention discloses a pathological feature recognition and negative elimination method based on microscopic imaging. The method comprises the following steps: S1, collecting a pathological section image and digitally generating original microscopic image data; s2, preprocessing the original microscopic image; s3, constructing a pathological image recognition network fusing converter coding and a gating dynamic receptive field mechanism, and outputting pathological feature vectors; s4, performing context modeling through an attention guidance and category perception decoder, and outputting an image classification result; s5, constructing a discriminant boundary separation model based on positive and negative sample embedding, and performing negative exclusion judgment; s6, performing confidence coefficient weighted evaluation in combination with the uncertainty and the boundary distance, setting a dynamic threshold value, and screening out low-credibility samples; and S7, coding the classification result and the negative label into structured data, and sending the structured data to a diagnosis auxiliary system. According to the method, multi-scale modeling and a negative screening mechanism are fused, and intelligent recognition and credible diagnosis output of the pathological image are realized.

The invention provides a microscopic image representation method based on a dynamic pluggable mask self-supervision encoder. The microscopic image representation method comprises the following steps: step 1, establishing a network based on the dynamic pluggable mask self-supervision encoder and an MAE decoder; 2, constructing a loss function to train the network, and carrying out the combined optimization of image reconstruction and classification; step 3, using a pre-training dynamic pluggable mask self-supervision encoder to extract deep feature representation of the microscopic image; and further connecting a decoder of a downstream task, and carrying out mineral microscopic image super-resolution reconstruction, inclusion automatic analysis, intelligent diamond cleanliness rating and general cell segmentation. According to the method, the feature extraction quality of the field with the data volume disadvantage is improved through the pluggable module, and the problem that most categories of cross-field data sets are unbalanced is effectively solved. Meanwhile, compared with the addition of branches, the pluggable module reduces the extra calculation overhead brought by the addition of a structure by 50%, and endows the model with extremely strong domain mobility.

The utility model discloses a micro-fluidic chip imaging positioning device, and belongs to the technical field of micro-fluidic chip positioning. According to the technical scheme, the system comprises a microcosmic illumination module, a microcosmic imaging module, a chip moving and macroscopic illumination module, a macroscopic imaging module and an optical path multiplexing module; the microscopic illumination module comprises a fly's-eye lens assembly, the fly's-eye lens assembly comprises two fly's-eye lenses, the optical axes of the two fly's-eye lenses are parallel to each other, and the focus of each lens array element in one fly's-eye lens coincides with the center of the lens array element at the corresponding position of the other fly's-eye lens. The micro-fluidic chip clamp is applied to the aspect of micro-fluidic chip imaging positioning, overcomes the defects caused by an existing micro-fluidic chip clamp, solves the technical problems of universality, automation and high precision of micro-fluidic chip positioning, and has the characteristics of rapidness, automation, high precision and universality.

The invention relates to the technical field of medical image processing and computer vision, in particular to a parasitic ovum microscopic image detection method and system based on polymorphic prior, and the method comprises the steps: obtaining a to-be-detected microscopic image, and carrying out the feature extraction of the to-be-detected microscopic image through a convolutional neural network, and obtaining an initial feature map; constructing a polymorphic convolution kernel library based on preset biological morphological characteristics of the parasitic ova; performing deep convolution and feature fusion operation on the initial feature map by using a polymorphic convolution kernel library to generate a space attention map; performing feature enhancement processing on the initial feature map by using the spatial attention map to obtain an enhanced feature map; and performing bounding box regression and category prediction on the enhanced feature map to obtain a parasitic ovum detection result. According to the method, morphological priori and attention mechanisms are introduced, so that the problems of egg form similarity, background interference and the like are solved, accurate and robust automatic detection is realized, and the clinical diagnosis efficiency is remarkably improved.

The invention provides a carbon fiber composite material microcrack image segmentation method based on a two-stage super-resolution dynamic attention network, and the method comprises the steps: collecting a microscopic image of a to-be-detected carbon fiber composite material, and carrying out the preprocessing of the image, so as to improve the discrimination between a crack and a background; scanning the preprocessed image by adopting a multi-scale sliding window, carrying out microcrack feature extraction in combination with morphological Top-hat transformation, and positioning a candidate region with microcracks; cutting an image corresponding to the candidate region into sub-images, inputting the sub-images into a pre-trained super-resolution reconstruction network, and amplifying the sub-images to 2-8 times of the original size to obtain a high-resolution candidate region image; inputting the super-resolution enhanced candidate region image into a pre-trained convolutional neural network model for pixel-level crack segmentation to obtain a segmentation result of cracks in the candidate region; and mapping the segmented micro-crack region back to an original image coordinate system, marking the position and shape of the crack on the original image, and outputting a final micro-crack segmentation result.

The invention discloses a probiotic packaging stability control method based on deep learning, and the method comprises the following steps: collecting multi-source data in a packaging process, and carrying out the preprocessing of the multi-source data; performing time sequence feature analysis on the standardized structured data set, and extracting multi-scale time sequence features; performing target segmentation and multi-dimensional structure feature extraction on the standardized microscopic image data set; feature fusion is carried out, a probiotic packaging stability feature space is constructed, and a stable feature vector is generated; inputting a time sequence prediction network, and carrying out time sequence modeling and prediction processing; performing control parameter adjustment and constraint optimization based on a difference value between a prediction result and stability reference data; and executing and updating the packaging control adjustment scheme according to feedback to complete a packaging control closed loop. According to the method, deep learning and multi-source data analysis are fused, intelligent prediction and adaptive control of the probiotic packaging process are realized, and the method has the advantages of high stability and high precision.

The invention discloses an all-optical three-dimensional scanning confocal fluorescent microscopic imaging device and an implementation method thereof. According to the invention, through a deep learning driven adaptive regulation and control method, CNN is adopted to process spatial distribution data and dynamically regulate and control a phase hologram and the light intensity and phase compensation of a laser, so that the focal point of exciting light is subjected to aberration-free axial displacement, a bidirectional parallel optical scanning track of a two-dimensional scanning system is optimized, and all-optical three-dimensional scanning is realized; lSTM and TCN are combined to obtain a time sequence dependency relationship, a laser light source, an adjustable diaphragm, an electric focus-adjustable lens and a photoelectric detector are integrally controlled, efficient synchronization and automatic operation and high-speed axial focusing adjustment are realized, errors and time sequence mismatch are eliminated, optical characteristics of different samples and environmental interference are automatically adapted, and high-quality imaging is kept. The robustness and the applicable scene range of the system are improved; the method is used for model biological embryo real-time tracking, intracellular signal molecule dynamic visualization, cell membrane protein migration and aggregation observation and intracellular organelle interaction tracking.

The invention relates to the technical field of farmland fertilization optimization, and discloses a farmland fertilizer application amount analysis and diagnosis method based on machine learning. The method comprises the following steps: firstly, acquiring near infrared spectrum data and surface microscopic image data of farmland soil, and generating a soil comprehensive risk area; analyzing the change gradient of the migration rate of the nitrogen element in the region to evaluate the offset state, and analyzing the interaction strength of the phosphorus element and the potassium element to evaluate the asymmetric coupling state of the phosphorus element and the potassium element; calculating the farmland fertilizer demand degree according to the evaluation result; analyzing the matching deviation between the operation behavior of the peasant household and the fertilizer demand degree in combination with historical fertilization operation records, and generating an operation deviation value; a fertilization optimization strategy is generated in combination with the fertilizer demand degree and the operation deviation value, and execution parameters of the fertilizer applicator are dynamically adjusted through a self-adaptive control algorithm. According to the method, the multi-dimensional information of the soil can be synthesized, the fertilizer demand condition is accurately evaluated, the fertilization strategy is optimized, and the fertilization operation is dynamically adapted.

The invention discloses a multi-mode image information fusion method based on deep learning, and relates to the technical field of image processing, and the method comprises the steps: obtaining image data of an optical microscopic mode and an electronic microscopic mode, and carrying out the spatial resolution alignment; a quantum entangled state light field coding method is adopted for the optical microscopic image data to generate quantum enhancement features, and topological persistent coherence analysis is performed on the electronic microscopic image data to extract multi-scale structural features; the features are mapped to a tensor space, cross-modal feature fusion is carried out through a tensor ring decomposition method of dynamic rank adjustment, a joint representation tensor is generated, physical constraint reconstruction is carried out on the joint representation tensor, and a super-resolution fusion image is generated. Through fusion of quantum enhancement features and multi-scale topological features, tensor ring decomposition of dynamic rank adjustment and Tucker decomposition dimension reduction compression, efficient fusion of high-resolution images is achieved, calculation redundancy is reduced, and processing efficiency is remarkably improved.

The invention relates to the cross technical field of artificial intelligence and mineralogy, and discloses an AI mineral automatic identification method based on an optical microphotograph, comprising the following steps: step S1: data acquisition and preprocessing; s2, data classification and labeling: constructing a rock microscopic image data set by adopting a two-stage labeling process; s3, building and training a deep learning model; s4, model optimization and quantitative feature learning; s5, developing a software platform; and S6, hardware system combination and full-process automation are carried out. In the aspect of operation convenience, a full-automatic slice scanning hardware system is constructed and combined with an intelligent software platform, so that full-process automatic operation from automatic rock slice scanning, automatic mineral analysis to automatic PDF report generation is realized, manual intervention is greatly reduced, the operation difficulty and complexity are reduced, and the working efficiency is improved. And even an operator without rich experience can easily complete complex mineral identification work.

The invention relates to an intelligent metallographic defect detection system and method based on improved YOLOv11, and the system achieves the automation of a whole process from the collection of a microscopic image of a metal material to the precise recognition of a defect through the integration of a metallographic image collection subsystem, a rotating frame marking and enhancing subsystem and an improved neural network processing subsystem. A rotating target detection mechanism is introduced to adapt to a tilt defect form, a progressive data enhancement strategy is designed to strengthen small target feature learning, and a replacement feature extraction module and an embedded attention mechanism are adopted to optimize a YOLOv11 network structure, so that the industrial real-time performance is ensured finally, meanwhile, the metallographic defect detection precision is improved, and the high-precision quality inspection requirement is met. The method not only solves the key technical bottlenecks of difficult inclined defect positioning, difficult small target detection, difficult model deployment, weak anti-interference capability and the like in the existing metallographic defect detection, but also realizes industrial-grade efficient and automatic metallographic defect detection, and has wide application prospects and popularization values.

The invention discloses a rock and ore microscopic image splicing method and system based on deep learning, and relates to the field of image processing and the technical field of microscopes, and the method comprises the steps: obtaining a local rock and ore microscopic image of a rock and ore slice, carrying out the preprocessing of the local rock and ore microscopic image, and carrying out the overlapping region coarse registration of the preprocessed local rock and ore microscopic image through a phase correlation method; based on an improved image feature detection model, basic features and description features in the local rock and ore microscopic image after coarse registration are extracted, and local image features of the local rock and ore microscopic image are obtained; performing feature matching on the local image features of the two groups of local rock and mineral microscopic images by using an image feature matching model to obtain a matching corresponding relation of the local image features; and based on an image fusion algorithm of a homography matrix and a partial differential equation, splicing and optimizing the local rock and ore microscopic images in combination with a matching corresponding relation of local image features, and generating a large-view-field rock and ore microscopic image. According to the method, the complex transformation between the images can be better processed.

The invention discloses a time domain compression femtosecond holographic microscopy reconstruction method based on spatial domain and frequency domain joint learning. Constructing a time domain compression femtosecond holographic microscopy system for carrying out multi-moment information time domain compression coding on a measured sample and generating a single-frame snapshot measurement graph; constructing a joint spatial domain-frequency domain learning end-to-end reconstruction model comprising a spatial frequency domain interaction module; training the model by taking a dynamic micro-nano structure hologram sequence generated by simulation as a training set until the joint spatial frequency domain loss function is converged; inputting a single-frame snapshot measurement image generated by a microscopic system into the model for processing to obtain a reconstructed hologram frame sequence; and performing frequency spectrum extraction and phase unwrapping processing to generate a dynamic three-dimensional scene video so as to realize time domain compression femtosecond holographic microscopy reconstruction. According to the invention, both the hardware imaging rate and the algorithm reconstruction precision are considered, and high-speed, high-resolution and low-cost four-dimensional holographic microscopic imaging can be realized under the conditions of extremely low data bandwidth and a common CMOS camera.

The invention discloses a method and system for recognizing the number of bacillus in a gynecological micro-ecological microscopic image, and belongs to the technical field of image recognition and micro-ecological analysis. Performing image enhancement processing on the image to improve the definition of the target area; pixel-level segmentation is carried out on the enhanced image based on the trained image segmentation model, and a suspected bacillus target area set is extracted; extracting a morphological characteristic parameter set for each target area; screening the suspected areas by combining a bacillus morphological feature discrimination model obtained by clinical labeling sample training, and removing false targets; counting the number of the effective areas, estimating the number of bacilli in the adhesion areas by adopting a skeleton endpoint analysis method, and finally outputting the total number of bacilli in the image; the method has the characteristics of high precision, high robustness and high automation degree, and is suitable for intelligent identification of the micro-ecological structure in the gynecological microscopic image.

The invention provides a deep learning-based junction point relation network training and cell occlusion judgment method, which comprises the following steps of: performing instance segmentation on a cell microscopic image to obtain a cell instance mask, and extracting a cell boundary based on the cell instance mask; performing junction point detection on the cell boundary to obtain cell shielding junction points, and classifying the cell shielding junction points into cell T-shaped shielding points, cell Y-shaped convergent points or cell X-shaped cross points; cutting the local image patch by taking the cell shielding junction point as a center, and constructing a multi-channel sample containing the local image patch; adopting a unified relationship label to label the multichannel sample, wherein the unified relationship label comprises a shielding relationship, a same-layer relationship and an unjudgeable relationship; training a deep learning network by using the labeled multi-channel sample, wherein the deep learning network outputs probability distribution of a junction point relationship category; and inputting a multi-channel sample of a to-be-detected cell microscopic image into the trained deep learning network, and outputting a junction point relationship category and a confidence coefficient thereof.

The invention relates to the technical field of digital core modeling and artificial intelligence modeling, in particular to a weathered sandstone microscopic image generation method based on a conditional generative adversarial network, which comprises the following steps: acquiring and analyzing sandstone samples with different weathering degrees to construct a weathered sandstone multi-scale database; performing weathering geological knowledge extraction on the weathered sandstone multi-scale database to construct a sandstone feature tag set; and training the conditional generative adversarial network by using the sandstone feature tag set to obtain a weathered sandstone microscopic image generation model, and generating a weathered sandstone microstructure image with physical consistency by using the weathered sandstone microscopic image generation model. Therefore, the problems that in existing weathered sandstone physical and mechanical property modeling, the depth incidence relation between the microstructure and the macro-mechanical property is difficult to reveal, and a traditional modeling method is difficult to effectively fuse the advantages of a weathering geological mechanism and modeling generation, so that a reconstruction result lacks physical consistency and weathering response rationality are solved.

The invention relates to the technical field of image processing, and discloses a fluorescent lifetime microscopic image large-view-field splicing method and system and electronic equipment, and the method comprises the steps: constructing a tissue region mask for a to-be-spliced image, and constructing a vignetting model in a tissue region to complete vignetting correction; counting brightness indexes in an organization area to determine a global brightness reference, and realizing image group brightness unification through global zooming and single image brightness adaptive correction; further realizing alignment of adjacent images through feature point matching, and finishing image block splicing in combination with a minimum color difference suture fusion method to obtain an image band; an overlapping area is extracted from an image belt to generate an effective content mask, relative displacement is estimated by using a phase correlation method after low-frequency suppression, image belt splicing is completed by multiplexing a minimum color difference suture fusion method, a large-view-field spliced image is output, the automation degree and robustness of splicing are improved, the spliced image is geometrically consistent and visually seamless, and the splicing efficiency is improved. And the requirements of medical research on high-resolution and large-field-of-view fluorescence lifetime microscopic images are met.

The invention relates to a sperm cell analysis and diagnosis system based on a multi-modal large language model, which comprises a multi-modal data co-processing unit, a cross-modal semantic alignment module, a dynamic diagnosis decision engine and a self-adaptive evolution system, the four-dimensional data processing module is used for synchronously processing microscopic images, motion trail videos, biochemical detection data and four-dimensional input data of medical record texts and comprises a feature selector based on a gating attention mechanism. According to the sperm cell analysis and diagnosis system based on the multi-modal large language model, quantitative analysis of sperm movement chaos features is realized for the first time, a nonlinear dynamic evaluation standard is established, a cross-modal knowledge distillation and meta-learning migration framework is developed, a data annotation dependence bottleneck is broken through, and an interpretable clinical decision support system is constructed; dynamic updating and probabilistic suggestion of diagnosis rules are achieved, semantic analysis of single-cell multi-omics data is achieved, and molecular mechanism research results are converted into clinically available knowledge.

The invention provides a comprehensive evaluation method for dissemination characteristics of minerals in coal of different particle sizes, and relates to the field of coal quality analysis. The method comprises the following steps: collecting coal analysis samples of different size fractions, and preparing coal polished sections of all the size fractions; the method comprises the following steps: collecting microscopic images, and constructing a representative microscopic image sample database of each size fraction; carrying out size fraction judgment, generating a component segmentation image on a single particle and a component segmentation image on a microscopic image, and extracting dissemination characteristic parameters of mineral substances in coal of each size fraction; key parameters are screened out, and comprehensive evaluation is carried out on the dissemination characteristics of the minerals in the coal of different size fractions. According to the method, an image processing technology and multivariate statistical analysis are combined, and qualitative and quantitative analysis of the dissemination characteristics of the minerals in the coal is achieved. Compared with an existing microscopic image analysis method, the method has the advantages that the human subjective influence and the labor intensity are remarkably reduced, and the analysis result has higher objectivity and reproducibility.

The invention relates to the field of image processing, and provides a seed multi-modal microscopic image acquisition method, device and equipment and a storage medium, and the method comprises the steps: carrying out the region segmentation of an obtained hyperspectral image of a plurality of seeds, and obtaining a single-seed region image; performing data extraction on the single seed area image to obtain spectral data of each seed; the spectral data comprises a seed identifier and a reflectivity value of each wave band; segmenting and extracting the acquired scanning images of the multiple seeds to obtain single seed scanning slices; performing three-dimensional reconstruction on the binary mask image to obtain seed three-dimensional physical structure information; the binary mask image is obtained by performing pixel classification processing on the single seed scanning slice. The method provides a basis for comprehensively and deeply analyzing chemical components and internal structures of the seeds, and improves the efficiency of seed detection and analysis.

A deep learning-based super-resolution fluorescence lifetime imaging microscopy (SR-FLIM) method includes the steps of: S1, performing fluorescence microscopic imaging on a sample to obtain confocal intensity images and stimulated emission depletion (STED) intensity images at a same location; S2, co-registering the acquired confocal and STED intensity images; S3, pairing the co-registered confocal and STED intensity images as input (Input) and ground truth (GT) to assemble a dataset; S4, partitioning the dataset into training and validation sets following a predefined ratio; and S5, constructing a network, and selecting hyperparameters and an optimizer. This method may achieve SR-FLIM within a conventional confocal FLIM system, surpassing spatial resolution limitations of FLIM, breaking through resolution barriers of conventional optical microscopy, while preserving normal fluorescence lifetime characteristics of fluorescent probes.

The invention provides a super-resolution fluorescence microscopic imaging method and product based on deep learning, and the method comprises the steps: carrying out the feature extraction and image reconstruction of an actual fluorescence microscopic image based on a trained super-resolution fluorescence microscopic imaging model, and obtaining a super-resolution fluorescence microscopic image result; wherein the super-resolution fluorescence microscopic imaging model is constructed on the basis of an encoder-decoder architecture fused with a residual network, a channel attention mechanism and a self-attention module, and the trained super-resolution fluorescence microscopic imaging model is obtained through pre-training-fine tuning dual-stage training. According to the invention, based on the encoder-decoder architecture, the residual network, the channel attention mechanism and the self-attention module construction are fused, and the pre-training and fine-tuning dual-stage training strategy is combined, so that the cross-modal super-resolution migration capability under the single-sample condition is realized, the bottleneck that the traditional deep learning depends on a large amount of data repeated training is broken through, and the super-resolution migration capability is improved. When different imaging samples and systems are replaced for super-resolution reconstruction, resource consumption can be remarkably reduced.

The invention belongs to the technical field of geological sample analysis, and particularly relates to a heavy mineral fidelity pretreatment and morphology automatic characterization method based on artificial intelligence. The method is based on a medium-low hardness heavy mineral fidelity separation and purification technology, an integrated multifunctional target manufacturing technology, a microscopic image intelligent acquisition technology and intelligent identification and feature extraction based on deep learning. The method comprises the following steps: acquiring fidelity heavy mineral particles which maintain the original three-dimensional morphology, are not reset in internal thermosensitive geological records and are not subjected to any systematic deviation screening on population composition, calculating a series of morphological parameters such as an extension coefficient, roundness, color and the like of each heavy mineral particle, and finally outputting a structured quantitative data report. According to the method, multidirectional and intelligent fidelity is realized from the source, and the analysis flux and objectivity are greatly improved by an automatic process, so that systematic and quantitative mineral resource general survey on massive fine particles in a giant deposition system becomes possible.

The invention discloses a living cell super-resolution imaging method and system based on self-supervised learning. The method comprises the following steps: S1, obtaining a fluorescence microscopic image through a standard fluorescence microscopic system; s2, for the acquired single noise image, generating positive and negative sample pairs of self-supervised training data through an autonomously designed and optimized self-supervised strategy; s3, constructing a front denoising neural network and a rear deconvolution network, and performing network training; and S4, inputting a newly obtained noise image into the trained front denoising network to obtain a denoised image, and inputting the denoised image into the rear deconvolution network to obtain a final super-resolution reconstructed image. According to the invention, the standard fluorescence microscopic system and the self-supervised denoising processing module are combined, so that high-quality denoising and super-resolution reconstruction of the fluorescence microscopic image can be realized in a low-photon signal scene.

A frequency-spatial domain alignment based method for super-resolution processing for a microscopic image time series. The method comprises: providing a low-resolution fluorescence microscopic image sequence, which consists of P low-resolution fluorescence microscopic images, for a biological sample, wherein P is a positive integer greater than 3; and feeding the low-resolution fluorescence microscopic image sequence into a super-resolution processing system consisting of a feature extraction module (220), a feature propagation and alignment module (230) and a reconstruction module (240).

The invention discloses a method for measuring the original austenite grain size of medium carbon steel. The method comprises the following steps: heating a medium carbon steel sample to a temperature above an original austenitizing temperature, and keeping the temperature for a period of time to precipitate carbides from an original austenite grain boundary; cutting and embedding the heat-treated medium carbon steel sample to prepare a medium carbon steel metallographic sample; carrying out grinding and fine polishing treatment on the medium carbon steel metallographic sample, and carrying out ultrasonic cleaning on the medium carbon steel metallographic sample in alcohol; immersing and corroding for a certain time by using nitric acid alcohol, then cleaning the surface of the medium carbon steel metallographic sample by using alcohol, and drying until an austenite grain boundary is clearly displayed; completing microscopic image rating; and calculating the austenite grain grade by adopting an intercept point method. According to the method, the original austenite grain boundary can be clearly displayed, and the effective size of the original austenite grain size of the medium carbon steel can be rapidly and accurately obtained.

The invention belongs to the technical field of image analysis, and particularly discloses an acute leukemia classification method and system based on frequency domain attention and multi-scale fusion, and the method comprises the following steps: collecting an acute leukemia microscopic image of a patient, inputting the acute leukemia microscopic image into a deep residual network ResNet50 backbone network, a frequency domain attention module DCT is introduced into a plurality of deep processing stages, an efficient multi-scale attention module EMA is introduced between the last-stage deep processing stage and the frequency domain attention module, and a subnet is remodeled through channel grouping; the output ends of all the frequency domain attention modules are input into the multi-scale feature pyramid network, and a fusion feature set is extracted; features extracted by the backbone network and the multi-scale feature pyramid network are fused, a full-connection classification head is input, and a classification result is output. By the adoption of the technical scheme, the problems that a traditional network is weak in tiny feature extraction capacity and a model is insufficient in multi-scale feature fusion are solved, and a more accurate classification result is obtained.

The invention discloses a coke microscopic optical tissue extraction method based on a multi-attention neural network, and the method comprises the following steps: obtaining and cutting a coke microscopic image, and constructing a data set containing a manually labeled optical tissue label and a background label; dividing the data set into a training set, a verification set and a test set, and performing data enhancement on the training set; constructing a multi-attention U-shaped neural network model, wherein the multi-attention U-shaped neural network model comprises an encoder, a bridging module, a bottleneck module and a decoder which are connected in sequence; a training set is adopted to train the multi-attention U-type neural network model, multi-scale prediction masks are generated at all stages of a decoder through a deep supervision mechanism, and the multi-scale prediction masks and final output jointly participate in loss function calculation; and inputting a to-be-processed coke microscopic image into the trained multi-attention U-shaped neural network model, and outputting an optical tissue extraction result.

The invention provides a lithium ore microscopic image segmentation method and system based on an improved Unet model, and belongs to the field of machine vision and image segmentation. The method comprises the following steps: constructing a lithium ore microscopic image data set, obtaining a microscopic image of a lithium ore sample to be detected, and preprocessing the image; pixel-level labeling is carried out on the microscopic image data set, a mask image corresponding to a mineral category is generated, and division and data enhancement are carried out; a microscopic image segmentation model is constructed and improved based on a Unet algorithm, a cascade structure of deep convolution and point-by-point convolution is constructed in a down-sampling path, an anti-adhesion dynamic snakelike convolution module is embedded in jump connection, and a channel-space dual-path adaptive attention mechanism module is introduced at the same time. A multi-scale feature aggregation module is constructed in the up-sampling path; a mature model is obtained after training; a to-be-detected microscopic image is input to obtain a preliminary segmentation image, and then adaptive area filtering is carried out. According to the invention, the segmentation accuracy and segmentation precision of the lithium ore microscopic image are improved.