136 results about "Image segmentation algorithm" patented technology

The invention discloses a plant extraction method and system based on image processing, and relates to the technical field of image processing. The method comprises the following steps: acquiring and standardizing a plant raw material image; extracting feature vectors of the images by using a neural network convolutional layer algorithm, comparing the feature vectors with a plant raw material standard feature database, and screening out qualified images of the plant raw materials; qualified images are processed through multispectral image analysis and an image segmentation algorithm, and an effective component spatial distribution diagram is obtained; constructing a partial least squares regression linear model, extracting spectral feature vectors of the effective components, inputting the spectral feature vectors into the model to obtain concentration values, and calculating quantitative data of the effective components; establishing an extraction process rule base, selecting a process parameter combination according to a spatial distribution diagram and quantitative data, and extracting to obtain a primary extracting solution and plant residues; and calculating an effective component residual error rate by matching and mapping to a qualified image coordinate, and if the effective component residual error rate is greater than a preset threshold, performing secondary extraction to obtain a secondary extracting solution. And quantitative basis is provided for extracting process parameters through the spatial distribution diagram of the effective components.

The invention discloses an agricultural insurance data acquisition and management system based on satellite remote sensing. The system comprises a data acquisition module, a preprocessing module, a verification module and a management module. The data acquisition module covers basic, universal and vector data acquisition according to data types; the preprocessing module improves the quality of collected data; the verification module verifies a cultivated land vector and insured user information, and data accuracy and reliability are enhanced; the management module builds two libraries, and constructs an agricultural insurance data large model by utilizing knowledge graph modeling. In the technical application, on the basis of satellite remote sensing and A I, large-user plot data is collected by means of an image segmentation algorithm, and the cost is reduced. A full-coverage model is constructed through abstract materialization, and business application is supported. Technical means such as satellite remote sensing, weight dispatch data superposition and AI image recognition are used for verifying data, manual operation is reduced, verification accuracy is guaranteed, data value is mined, and the agricultural insurance data management level is improved.

The invention provides an image processing system and device of an intelligent water affair inspection robot, relates to the technical field of water affair inspection, and solves the technical problems of poor light reflection restoration effect, difficulty in floating object feature extraction and relatively high water quality abnormity judgment misjudgment rate in the prior art. The system comprises an image processing module, and an acquisition module and an image analysis module which are connected with the image processing module, the acquisition module is used for acquiring an original image of a water body area acquired by the inspection robot; the image processing module is used for performing reflection restoration on a reflection area in the original image through pixel restoration to obtain a restored image; performing water body and surrounding environment segmentation on the restored image through an image segmentation algorithm to obtain a water body image; performing feature extraction on the water body image through edge detection to obtain water quality features; and the image analysis module is used for carrying out abnormity judgment on the water body area according to the water quality characteristics and generating an abnormal signal. The method is used in the water affair inspection process.

Systems and methods for detecting artificial intelligence generated images are provided. The system accepts an input image (e.g., a digital still image, or a frame from a digital video file or image stream) and subdivides the input image into a set of patches using a patch partitioning algorithm. The system then processes each patch and produces a feature embedding for each patch within a high dimension space. The system then utilizes these patches with further processing as input to machine learning models, which allows the system to achieve image, patch-level, and video-frame generated image classification and localization alongside identification of the generative model used to synthesize the image.

The invention, which belongs to the technical field of medical image processing and computer vision, discloses a multispectral imaging-assisted post-radiotherapy skin injury early detection system comprising a multispectral data acquisition module, a deep tissue feature extraction module, a three-dimensional lesion segmentation module and a space-time tracking evaluation module. The subcutaneous 2.5 cm depth tissue information is obtained through multispectral imaging at the wave band of 400-1350nm, blood perfusion, melanin concentration, collagen structure and other physiological parameters are inversed based on the radiation transfer theory, a three-dimensional medical image segmentation algorithm is adopted to achieve three-dimensional accurate segmentation of an injury area, a spatio-temporal evolution model is established to predict the injury development trend, and the damage development trend is predicted. The radioactive skin injury can be detected in the subclinical period, the detection time window is advanced by 5.2 days on average, the occurrence rate of severe dermatitis is reduced by 65%, and a basis is provided for clinical timely intervention.

The radar target scattering center region segmentation method based on local density clustering first carries out Frost filtering to the original radar image to realize global noise suppression of the image; secondly, the image segmentation is carried out to the Frost filtered image based on the level set method, and the suspected target region of interest is preliminarily obtained; then, the area filtering is adopted to extract the maximum connected domain, and the target region of interest is obtained; finally, the scattering center is detected and the region segmentation is carried out by using the local density clustering algorithm. The radar target scattering center region segmentation method based on local density clustering has high accuracy, and can effectively improve the technical problem that the SAR image is difficult to have high enough resolution in the actual situation, so that the traditional image segmentation algorithm is difficult to segment the region containing only one scattering center.

The invention relates to the technical field of alimentary canal diagnosis and treatment, in particular to an alimentary canal detection drug delivery system which comprises an alimentary canal sensing unit for collecting an image, a pH value, a temperature and pathological tissue pathological characteristic data in an alimentary canal cavity; the lesion positioning analysis unit identifies a lesion area through an image segmentation algorithm, constructs a lesion three-dimensional coordinate model according to physiological parameters, prejudges a lesion displacement trend in combination with an alimentary canal peristalsis rule, analyzes a lesion state in stages, and outputs lesion severity; the response type drug delivery module controls the micro drug delivery device to perform targeted release and adjust the drug release rate based on the lesion coordinates and the intracavity environment parameters; the data collaboration platform shares a detection result and a drug administration record through a real-time interaction interface; the curative effect feedback unit is used for collecting physiological index change and patient symptom improvement data after drug administration, judging the drug onset period and possible adverse reactions, and optimizing drug administration parameters and detection frequency. Therefore, the problems of low targeting precision, poor diagnosis and treatment efficiency and the like in the prior art are solved.

The invention discloses a curative effect evaluation system for treating liver cancer by combining HDRBT with thermal ablation in multi-modal radiomics, and relates to the field of liver cancer treatment, and the system comprises a multi-modal image data acquisition module which is used for collecting multi-modal image data; the three-dimensional visual model construction module is used for reconstructing a three-dimensional structure of liver tumors, peripheral blood vessels and bile ducts through an image segmentation algorithm, and generating a visual model by fusing radiomics characteristics; the combined treatment plan generation module is used for integrating the dosimetry parameters of the HDRBT and the process parameters of the thermal ablation, optimizing a catheter implanting path, the position of a thermal ablation electrode and a treatment sequence, and generating a combined treatment scheme; and the curative effect evaluation module is used for quantitatively analyzing tumor volume change, ablation boundary integrity and residual tumor activity through accurate registration of multi-modal images before and after treatment. According to the scheme, the problems of incomplete thermal ablation and high recurrence rate of hepatocellular carcinoma can be solved, and full-chain technical support is provided for clinical precise application.

The application discloses an image segmentation method and system for double-layer liquid crystal screen display, the double-layer liquid crystal screen is a front liquid crystal screen close to an audience and a rear liquid crystal screen close to a backlight module, and the method comprises the following steps: constructing a convolutional neural network model, the convolutional neural network model is used for processing an input image to obtain a first image and a second image, and the second image is sent to the rear liquid crystal screen, wherein the first image is used for being sent to the front liquid crystal screen; through the first image, the second image is subjected to image reconstruction, the image reconstruction is used for improving the artifact problem of double-screen liquid crystal display and improving the image display quality; and the second image after reconstruction is sent to the rear liquid crystal screen; the application has great optimization effects on the artifact phenomenon and the image quality of the double-layer screen, and the calculation speed is greatly improved compared with an image segmentation algorithm based on viewpoint compensation.

Disclosed herein is a medical system (100, 300) comprising a memory (110) storing machine executable instructions (120) and an image segmentation algorithm (122). The image segmentation algorithm is configured for outputting one or more predetermined anatomical regions within initial magnetic resonance imaging data (124) descriptive of a predetermined field of view (109) of a subject (318). The medical system further comprises a computational system (104), wherein execution of the machine executable instructions causes the computational system to: receive (200) the initial magnetic resonance imaging data (124); receive (202) the image segmentation comprising the one or more anatomical regions within the magnetic resonance imaging data in response to inputting the initial magnetic resonance imaging data into the image segmentation algorithm; select (204) at least one of the one or more anatomical regions as a selected image portion (128) using a predetermined criterion; and reduce (206) image intensity within the selected image portion to provide intensity corrected magnetic resonance imaging data.

The present application provides a kind of geologic stratum lithology regional prediction method, system, terminal equipment and storage medium, it is related to computer system field.The system includes data management module, sample management module, model training module, model management module, lithology prediction module and the like.The present application utilizes geophysical, geochemical, remote sensing and other multi-source data, based on deep learning image segmentation algorithm, forms regional shallow surface stratum lithology identification model, carries out high-precision rapid stratum lithology prediction and regional segmentation to the bedrock exposed area and shallow cover area with lower exploration degree, supports geological mapping and mineral exploration work.

Disclosed is an image-based lost ball automatic classification method comprising: a learning image generation step of generating a brand learning image for learning a brand of a golf ball and a defect learning image for learning a defect; a golf ball brand learning step of training an image classification algorithm about the brand of the golf ball by using the brand learning image; a golf ball defect learning step of training an image segmentation algorithm about the defect in the golf ball by using the defect learning image; a target image generation step of generating at least eight target images from a target lost ball to be classified; a target lost ball brand classification step of classifying a brand of the target lost ball from a target image by using the trained image classification algorithm; and a target lost ball grade classification step of classifying a grade of the target lost ball from the target image by using the trained image segmentation algorithm.

The invention relates to a multispectral image optical performance detection method, system and device and a storage medium, and relates to the field of image optical performance detection. The method comprises the following steps: acquiring a multispectral image containing a plurality of cross targets by using a high-resolution camera; extracting and separating the multispectral image according to the plurality of cross targets to obtain a plurality of target multispectral images, and transmitting the plurality of target multispectral images to an image detection and analysis module; when the image detection and analysis module receives a target multispectral image, the target multispectral image is segmented by using an adaptive Niblack image segmentation algorithm to obtain a binary multispectral image; an improved double-iteration parallel skeleton extraction algorithm is adopted to extract a cross target skeleton of the binarized multispectral image; adopting a neighborhood analysis method to extract center coordinates of the cross target and storing center coordinate data; and calculating the optical performance index of the multispectral image according to the plurality of pieces of center coordinate data. The technical effect of the invention is that the optical performance of the multispectral image can be detected more efficiently.

The invention provides a crane anti-collision method and device based on visual identification, relates to the technical field of computer vision, and solves the technical problem that the anti-collision requirements of high precision and high real-time performance cannot be met in the prior art. The method comprises the following steps: acquiring video stream data through video acquisition equipment; performing image segmentation on the video stream data through an image segmentation algorithm to obtain a hoisting box contour and a stacking contour; calculating a minimum spacing sequence between the lifting box and the stack in the plurality of video frames based on the contour of the lifting box and the contour of the stack; calculating the swinging characteristics of the hoisting box based on the minimum spacing sequence; according to the swinging characteristics, the swinging amplitude of the hoisting box is predicted, and then the predicted minimum distance between the hoisting box and the stack is calculated; and evaluating a collision risk level according to the predicted minimum spacing, and generating a control instruction. The anti-collision device is used in the anti-collision process of the crane.

The application discloses a kind of modulation signal envelope extraction algorithm and electronic equipment, wherein algorithm includes the following steps: S1, according to modulation data, generate waveform image and obtain the image area corresponding to waveform image;S2, extract the original boundary of image area, obtain boundary pixel coordinate point set;S3, the coordinate point set after screening processing is handled to boundary pixel;S4, based on the coordinate point set after screening processing, complete envelope extraction, obtain the envelope set of modulation signal.The present application converts discrete data into waveform image, extracts target area using image segmentation algorithm, extracts area edge point coordinates using edge extraction algorithm of image processing, and then converts irregular area boundary at pixel level into an ordered coordinate set after screening, reduces data volume and improves the accuracy of subsequent envelope extraction, effectively solves the shortcomings in the existing commonly used algorithm;The present application has the characteristics of fast detection speed, safety and reliability, good practicability, high accuracy and the like.

The invention relates to a perception identification method for verification and evaluation of a space autonomous capture full task chain, and the method comprises the steps: shooting a space region containing a target through a fixed binocular camera, and reconstructing a three-dimensional point cloud model of the space region where the target is located; a mask set of a target to be grabbed is screened out through a deep learning image segmentation algorithm and target feature comparison, a target three-dimensional point cloud is obtained through mapping, geometric centroid calculation and coordinate system conversion are carried out on the point cloud, and a three-dimensional center coordinate of the target under a base coordinate system of the space robot is obtained; a monocular camera carried at the tail end of the space robot is used for shooting multi-view images of a target, and a high-precision target three-dimensional point cloud model with a continuous surface is generated; carrying out fusion registration on the high-precision target three-dimensional point cloud model and the three-dimensional point cloud model of the space region where the target is located to obtain a target complete geometric model with accurate scale and pose; and the optimal grabbing pose and path are calculated. And the problem of autonomous perception and recognition of the robot in the autonomous grabbing process in the space environment is solved.

The application discloses a tree trunk direction extraction method and system based on layered symmetry analysis, which comprises the following steps: preprocessing a tree point cloud; based on a crown elevation model of the point cloud, using an image segmentation algorithm, the preprocessed point cloud is segmented into single-tree point clouds; the single-tree point clouds are divided into multiple non-equidistant height layers in the Z-axis direction; the most symmetric direction of the point cloud in each height layer is found; the angle between the most symmetric direction and the average direction of each height layer is calculated, and the most symmetric direction with an angle greater than a set threshold is removed; and the most symmetric direction of each height layer is weighted and averaged to obtain the trunk direction of the tree. The application uses a method based on layered symmetry analysis to identify the main direction, which is more in line with the natural posture of the tree, thereby improving the fitting accuracy of the trunk direction and having good robustness and generalization ability under complex forest environment and noise interference.

The application discloses a low-cost, high-precision and high-reliability method for unmanned ship intelligent environment perception by fusing image segmentation and millimeter wave radar, which can acquire the environment information of the route and the surrounding of the unmanned ship in real time during the driving of the unmanned ship, including the shoreline, the sailing ship, the navigation mark and the floating object, and provide decision support for the real-time route adjustment and obstacle avoidance of the unmanned ship. The method comprises four parts of the millimeter wave radar, the binocular camera, the image segmentation module and the information fusion module, the image segmentation module is responsible for reading the image data, executing the image segmentation algorithm, identifying the target and generating the minimum fitting rectangle of the target, the information fusion module is responsible for receiving the data of the radar and the image segmentation module, acquiring the polar coordinates of the fusion target, matching the target size and outputting the result to the controller.

The application provides a guide wire bending energy calculation method, an interventional robot and a storage medium. The method comprises the following steps: acquiring a guide wire tip image and inputting the guide wire tip image into a convolution module to obtain a first feature image; performing image segmentation on the first feature image according to an image segmentation algorithm to obtain a plurality of feature image blocks; mapping the plurality of feature image blocks to an embedding layer to obtain a second feature image; inputting the second feature image into a Transformer module to obtain a third feature image; performing image connection on the first feature image and the third feature image to obtain a target feature image, and inputting the target feature image into a segmentation head module for image segmentation processing to obtain a guide wire segmentation image; and obtaining a guide wire bending energy calculation result according to imaging state information of the guide wire segmentation image and the guide wire tip image. The application calculates the guide wire bending energy in real time, and replaces the doctor in the operation in the radiation environment through the interventional robot, so that the operation stability of the doctor is ensured and the risk of the operation is reduced.

The application discloses a kind of based on deep learning multi-modal medical image segmentation method, the method includes the following steps: one, obtains multi-modal medical image and standardizes size;Two, modality 1-3 are input into the FCSA fusion module of the channel 2 of segmentation network and obtain the feature map after fusion;Three, leading modality 4 is input into the channel 1 of segmentation network;Four, the channel 1 and channel 2 of segmentation network are all extracted by five layers improved Tok-KAN module and down sampling to different scale feature information;Five, in decoding process, the different scale encoding feature map of each layer output is recovered and decoded by the decoder of dense jump connection;Six, by upsampling and Sigmoid activation, generate segmentation prediction from the output of the last feature extractor of decoder.The application solves the problems of low segmentation accuracy and poor model interpretability of multi-modal medical image segmentation algorithm.

The application discloses a wall crack visual detection method based on an adaptive image segmentation algorithm, and the method comprises the following steps: performing a gray-scale conversion on an original wall crack image, taking a complete gray value range as a search interval, performing a morphological operation and a filter pretreatment on the gray-scale image, traversing each threshold value in the interval to binarize the pretreated image, and extracting a morphological topology and a geometric feature sequence corresponding to each threshold value; determining an effective threshold value search space by taking a point with the largest number of connected domains as a lower limit, performing linear normalization on the morphological topology feature sequence, and performing logarithmic smoothing and normalization processing on the geometric feature sequence; constructing a coupled double-target fitness evaluation function to obtain an optimal segmentation threshold value; obtaining a binarization result by using the optimal threshold value, removing small-area noise points and connecting broken cracks through a morphological operation, and obtaining a final crack binarization image. The method has high recognition accuracy, can adapt to uneven illumination and a large noise environment, and has a good engineering application prospect.

The application discloses a ship lock concrete structure durability analysis method and system under the coupling action of multiple factors, which comprises the following steps: acquiring internal structure images of different erosion areas of a ship lock concrete test piece; identifying and quantitatively analyzing the internal structure images of different erosion areas by a deep learning image segmentation algorithm to generate an erosion structure factor set; calculating environmental characteristics corresponding to multiple environmental parameters by a multi-physical field coupling inversion algorithm; constructing a three-dimensional microstructure model of different erosion areas; dynamically simulating the three-dimensional microstructure model according to the environmental characteristics; performing spatio-temporal correlation analysis on simulation result data by a hypergraph neural network to generate a prediction result; and comparing an expert knowledge base and performing construction maintenance based on text information of the prediction result by an Euclidean distance similarity algorithm. The application introduces high-precision three-dimensional model multi-factor coupling simulation, and improves the ship lock concrete structure analysis precision through high-precision coupling prediction of the hypergraph neural network.

This application belongs to the field of image processing technology and discloses a method and apparatus for generating hairstyle replacement videos based on a diffusion model. The method includes: acquiring a hairstyle image dataset and training a diffusion model to obtain a hairstyle generation model; acquiring a user-inputted image of a person and a target hairstyle prompt; using skeletal keypoint detection technology to identify the person image and obtain skeletal pose data; using an image segmentation algorithm to divide the hair region in the person image to obtain a hair mask; performing smoothing and feathering processing on the hair mask using Gaussian blur and feathering algorithms; inputting the hair mask and the target hairstyle prompt into the hairstyle generation model, and using the skeletal pose data as constraints for the hairstyle generation model to obtain a hairstyle replacement image; and generating a hairstyle replacement video based on the person image and the hairstyle replacement image. This application can improve the realism and naturalness of hairstyle replacement images and videos.

The invention provides a cerebral cortex thickness measurement method and system based on a level set image segmentation algorithm, and belongs to the technical field of neural image analysis, and the method comprises the steps: obtaining a brain magnetic resonance imaging image, and carrying out the skull stripping processing; performing three-dimensional visualization on the stripped image through volume rendering, and acquiring three-dimensional coordinates of points of interest and a region of interest on the surface of the cortex by using a preset interaction means; using a level set method to carry out brain grey matter and brain white matter segmentation on the region of interest to obtain a brain grey matter segmentation layer and a brain white matter segmentation layer, and storing three-dimensional coordinate data of points in the segmentation layers; calculating the minimum Euclidean distance according to the three-dimensional coordinate data of the points in the segmentation layer and the three-dimensional coordinates of the points of interest to obtain the cortex thickness of the points of interest; the method solves the problems that the existing cerebral cortex thickness measurement lacks a visual means, is not supported by a mathematical model, cannot be accurate to the thickness of an interest point, and cannot display and retain comprehensive information.

The invention discloses an underwater obstacle segmentation method and system based on unmanned ship side scanning, and relates to the technical field of underwater object detection. The method comprises the following three steps of underwater side-scan XTF file analysis, underwater side-scan waterfall plot denoising and underwater side-scan denoised image segmentation. According to the technical scheme, by means of optimization algorithms such as grid parameter adjustment and the like and PSNR scores, suitable selection parameters during BM3D denoising and graph segmentation based on QPSO + morphological parameter adjustment can be automatically obtained, compared with a traditional image segmentation algorithm, morphological segmentation can be carried out more quickly and more accurately through a threshold value, a clear and complete defect area contour can be obtained, and the defect area denoising accuracy is improved. The obstacle can be rapidly and accurately cut out, accurate positioning of underwater obstacles is achieved, and the requirements of an unmanned ship for seabed target searching and handling work are met.

The present application relates to the technical field of deep learning machine vision control convex letter steel rolling, and relates to a convex letter steel rolling plate shape control method and system based on deep learning, which comprises the following steps: heating and descaling a blank, setting a first visual detection unit at the outlet of a heating furnace, triggering hydraulic correction when the centering deviation is greater than a first threshold value; setting a second visual detection unit at a rough rolling mill, extracting the residual area of the rolling piece surface oxide scale through an image segmentation algorithm, and detecting whether to trigger secondary descaling; setting a third visual detection unit at a medium rolling mill, adjusting the flow distribution of the cooling water of a finishing mill if the temperature distribution is uneven, and correcting the side pressure of the finishing mill vertical roll if the corner radius error exceeds a fourth threshold value; setting a fourth visual detection unit at the finishing mill, calculating the roll gap compensation amount and correcting the finishing mill reduction amount in real time when the filling rate is less than a fifth threshold value. The present application combines convex letter steel with deep learning and machine vision, and can make the quality of the manufactured convex letter steel better and the shape more accurate.

A background replacement method and device, electronic equipment and computer readable storage medium, wherein the background replacement method comprises: obtaining a to-be-processed image; performing image segmentation on the to-be-processed image through an AI image segmentation algorithm to obtain a first matting result, the first matting result comprising first mask values of each pixel point in the to-be-processed image; performing color difference-based matting processing on the to-be-processed image through a color difference matting algorithm to obtain a second matting result, the second matting result comprising second mask values of each pixel point in the to-be-processed image; determining target mask values of each pixel point in the to-be-processed image according to the first mask values and the second mask values; and performing background replacement on the to-be-processed image according to the target mask values and a target background image to obtain a replacement image. The above background replacement method, device, electronic equipment and computer readable storage medium can achieve the purpose of improving the accuracy of background replacement.

The invention relates to a glass fiber diameter analysis method based on a microscopic image, and the method comprises the steps: giving a fuzzy edge through an image segmentation algorithm, marking out a rough diameter, and obtaining a glass fiber diameter near the rough diameter of an original microscopic image through the pixel value change rule characteristics of the glass fiber diameter on the microscopic image; and finding a left boundary and a right boundary of the accurate diameter on a derivative curve of the CubicSpline fitting curve. According to the method, the microscopic image is analyzed through an algorithm, errors generated when the diameter of the glass fiber is measured through manual scribing can be eliminated, final measurement is carried out on a derivative curve of a fine fitting curve, and the measurement accuracy is improved. In addition, by means of the method, automation of glass fiber diameter measurement can be achieved, the measurement efficiency is effectively improved, and high-flux accurate measurement of the glass fiber diameter becomes possible.

The application discloses a PIN needle position degree detection method and system, the method comprises the following steps: collecting a front image and a side image of the PIN needle, establishing a unified pixel coordinate and physical coordinate mapping relationship through joint calibration of the front camera and the side camera, and realizing spatial alignment of the front image and the side image; carrying out denoising and illumination equalization processing on the front image and the side image, and separating the PIN needle foreground area and the background area in the front image and the side image through an image segmentation algorithm; carrying out edge detection and sub-pixel interpolation processing on the separated image, and extracting the center line and the end point sub-pixel level coordinates of the PIN needle in the front image; calculating the position degree parameters of the PIN needle based on the extracted sub-pixel level features; comparing the calculated position degree parameters with the respective corresponding preset tolerance threshold, judging whether the PIN needle or the product corresponding thereto is qualified or not, outputting a detection result, and controlling a production line to perform corresponding actions according to the judgment result.

The present application belongs to the technical field of building material microstructure analysis, and in particular to a foam light soil microstructure analysis method based on image recognition; a foam light soil forming test piece is selected, and a flaky analysis sample is obtained after surface treatment; an imaging device is used to collect multi-view images of the sample, and two-dimensional images meeting the analysis accuracy are obtained; the collected images are subjected to denoising, segmentation and quality optimization processing, and the boundary features of the pores and matrix are highlighted; the pore area is extracted through an image segmentation algorithm, the pore profile is automatically identified and microstructure characteristic parameters are collected, and an analysis model is constructed based on the characteristic parameters; the present application includes sample preparation, multi-view image acquisition, optimization preprocessing, feature identification and analysis steps, and through multi-angle collection and algorithm optimization, the pore features are accurately quantified, the information loss and quantization errors of traditional methods are avoided, and the present application has the advantages of low cost, high efficiency and accurate analysis of foam light soil microstructure features.