100 results about "Region growing algorithm" patented technology

The invention relates to the technical field of remote sensing geological prediction, in particular to a slope support stability prediction method and system based on remote sensing data, and the method comprises the steps: obtaining a multispectral remote sensing image of a target slope region, and generating a vegetation mask through employing a normalized vegetation index; removing interference pixels of a vegetation coverage area in combination with morphological filtering and connected area analysis, then identifying support structure features through an improved edge detection algorithm, delimiting an influence area by adopting a region growing algorithm based on machine learning, inputting an area image into a convolutional neural network model, and obtaining an image of the support structure; a deformation probability graph is output by using a data enhancement technology and a weight optimization layer, finally, a threshold value is determined according to historical data, slope support stability categories are divided in combination with spatial neighborhood information and a voting mechanism, vegetation interference is effectively eliminated, a support structure is accurately recognized, prediction precision is improved, slope support stability can be accurately judged in time, and the method is suitable for popularization and application. And a reliable basis is provided for early warning and protection of slope disasters.

The invention relates to the technical field of vertical shaft guide detection, in particular to an industrial intelligent visual detection and diagnosis system for a vertical shaft guide. The coal dust shadow quantization unit extracts a coal dust region by adopting a double-peak self-adaptive threshold segmentation combined region growing algorithm, and introduces a space weight factor to calculate an average gray value and a coverage area proportion, and the dynamic coupling calibration unit depends on a three-dimensional lookup table calibrated by multiple working conditions and nonlinear interpolation; the optimal deviation compensation coefficient of the jitter and coal dust combination is matched, a dynamic gray segmentation threshold value is generated, and a defect diagnosis decision unit accurately calculates the abrasion depth through partition self-adaptive segmentation, multi-scale morphological filtering and denoising and gradient amplitude and curvature double-constrained sub-pixel edge detection. A hierarchical control instruction is output according to the real-time depth and the historical trend, coupling interference is eliminated, the cage guide abrasion detection accuracy is improved, and a reliable basis is provided for safe operation and maintenance.

The invention provides an intelligent labeling method for digital slices at a browser side and application of the intelligent labeling method, and relates to the technical field of medical image processing. The invention aims to solve the problem of poor interaction experience caused by data privacy risk and network delay due to dependence on a back-end server in the prior art. The method comprises the following steps: determining a seed point in response to an interactive operation of a user at a browser end, and dynamically loading a local slice image based on a spatial index; and calling a GPU (Graphics Processing Unit) by utilizing WebGL to calculate pixel similarity in parallel, adaptively adjusting a similarity threshold according to a current view zooming level, generating a segmentation mask in combination with a scanning line region growing algorithm, and performing vectorization rendering. According to the method, data does not need to be uploaded to a server, pure front-end, lightweight and real-time intelligent labeling of the GB-level high-resolution pathological section is achieved, data privacy is effectively guaranteed, and labeling efficiency is improved.

The invention provides a method, apparatus, and storage medium for reconstructing three-dimensional models of buildings based on missing point cloud data. The method includes integrating image-based point cloud generation, neural network techniques, and skeleton line extraction methods, offering a novel approach to handling missing point cloud data. The generation of point cloud data is achieved using principles of Structure from Motion based on video or panoramic image data. The point cloud is sampled and segmented using a region growing algorithm. A neural network based on PointNet is constructed, utilizing cross-entropy loss functions to assess the missing points in the point cloud. For mapping high-confidence point clouds from sampled points, Truth Points is employed to complete the entire process of real-world three-dimensional reconstruction. The integration of images into the three-dimensional scene is achieved with strict geometric relationships.

The invention relates to the technical field of image processing, and discloses a port loading and unloading risk identification method and system based on image processing. The method comprises the following steps: acquiring standardized image data through multispectral image acquisition and sea wind disturbance compensation processing, extracting multidimensional risk characteristics of suspension arm swinging, cargo deviation and personnel violation, performing dual evaluation of collision trajectory prediction and violation behavior detection, identifying comprehensive potential safety hazards by adopting a spatio-temporal correlation adaptive region growth algorithm, and determining whether the potential safety hazards exist or not. And generating port loading and unloading composite risk early warning information. The technical problems that multiple risk factors are difficult to accurately identify and the composite risk state cannot be effectively predicted in a complex marine environment in port loading and unloading operation are solved, and the environmental adaptability of port loading and unloading risk identification and the accuracy of composite risk assessment are remarkably improved.

The invention discloses a DBSCAN unmanned aerial vehicle path planning method based on depth estimation and optical flow analysis, and relates to the technical field of unmanned aerial vehicle autonomous navigation, and the method comprises the steps: extracting an image depth feature through a VGG16 model, carrying out the enhancement and segmentation of an image through Gamma transformation and a region growing algorithm, optimizing a seed group through optical flow analysis and a conditional random field algorithm, and carrying out the optimization of the depth feature of the image. A DBSCAN algorithm is used for carrying out path clustering, an A * algorithm is used for carrying out obstacle avoidance path planning, and a double-loop PID control algorithm is combined for execution. According to the invention, through fusion of multi-modal data and path clustering optimization, navigation precision and robustness in a complex environment are improved, seed group dynamic updating is realized based on seed probability calculation and optical flow analysis of a conditional random field, sensing precision, environmental adaptability and real-time reaction capability are improved, and through combination of double-loop PID control and an A * algorithm, the real-time response capability of the system is improved. And the real-time obstacle avoidance capability is optimized.

The invention provides a target identification method and device, electronic equipment and a storage medium. The method comprises the following steps: screening seed points according to a first distance between a second pixel point and a central point of a target mask area obtained by original point cloud projection and a depth value of an original point corresponding to the second pixel point; determining an adaptive search radius according to a second distance between the seed point and the original point corresponding to each second pixel point; carrying out adaptive region growth based on the adaptive seed point and the search radius to search the original point on the target; and determining a three-dimensional perception result of the target based on all the searched original points. According to the method, the adaptive seed point and the search radius can be obtained based on the characteristics of different targets, and an adaptive region growing algorithm is formed; therefore, the search requirements of different targets are met, and the target identification precision and speed are improved.

The invention discloses a steel pipe defect detection method based on artificial intelligence, and belongs to the technical field of pipe detection, and the method comprises the following steps: S1, obtaining original data from a steel pipe surface image, extracting initial features through a convolutional neural network, and generating a first feature set; s2, aiming at the first feature set, adopting an adaptive threshold segmentation algorithm to generate a second feature set; s3, according to the second feature set, performing preliminary division on the defect region through a region growing algorithm to obtain a plurality of defect region images; s4, extracting local texture features of each defect area image, and generating a third feature set corresponding to each defect area image; and S5, for the third feature set, performing defect type classification by adopting a pre-trained support vector machine classifier to obtain a defect type classification result. The steel pipe defect detection method based on artificial intelligence solves the problem that the precision and robustness of current steel pipe surface defect detection are difficult to improve.

The application provides an electrical equipment infrared image abnormality recognition method and system based on parallel operation characteristics, belongs to the technical field of image abnormality recognition, and acquires an infrared image of electrical equipment in parallel operation; an acquired infrared image is processed by using a pre-trained abnormality recognition model to obtain a recognition result of whether the electrical equipment is abnormal or not. The application utilizes a YOLO target detection algorithm to accurately segment key components of equipment, and then combines an improved region growing algorithm to position an abnormal temperature region in the infrared image. By fully utilizing the high consistency of two parallel equipment in electrical response and thermodynamic behavior, even under the condition of no fault sample, a local temperature abnormality region can be accurately recognized, and new technical support is provided for online state monitoring and intelligent operation and maintenance of electrical equipment.

The invention discloses a dangerous rock mass discontinuous edge point extraction method and device based on improved central axis transformation. The method comprises the following steps: acquiring discontinuous edge line candidate points of a rock mass from three-dimensional point cloud data with a normal in a target area; performing point cloud blocking on the three-dimensional point cloud data to obtain three-dimensional point cloud cluster data; calculating the curvature of discontinuous edge line candidate points in the selected three-dimensional point cloud cluster data, and separating the three-dimensional point cloud cluster data based on a region growing algorithm to obtain a separation region corresponding to each piece of three-dimensional point cloud cluster data; extracting three-dimensional point cloud cluster data and three-dimensional boundary point clouds of the partition areas; and fusing the three-dimensional point cloud cluster data and the three-dimensional boundary point clouds of the separated areas, and carrying out filtering to obtain a final dangerous rock mass discontinuous line. According to the method, the precision and reliability of discontinuous feature extraction of the dangerous rock mass are remarkably improved, the harsh requirement for the uniformity of initial data is reduced, and the robustness and universality under complex geological conditions and different acquisition environments are improved.

The invention discloses a segmentation modeling method for improving bronchial branch identification precision, and belongs to the technical field of medical image processing and respiratory tract three-dimensional reconstruction. According to the method, high-resolution acquisition is performed on a respiratory tract CT image, and preliminary segmentation of a tracheal tree is completed in combination with gray threshold setting and a region growing algorithm; and a reference line is introduced at the bifurcation, and dynamic region growth based on density gradient is adopted, so that refined identification of the sixth-level bronchial branch is realized. And then carrying out iteration correction on the generated model for multiple times, manually comparing anatomical features to delete pseudo branches and complement missing branches, and completing smoothing and NURBS curved surface construction in Geomagic software. And finally, splicing and fusing the bronchial branch model and an independently generated upper respiratory tract model to obtain a complete three-dimensional respiratory tract combination model, and ensuring that the error between the three-dimensional respiratory tract combination model and an original CT image is controlled within 0.8 mm through deviation analysis. According to the method, the recognition precision of the fine bronchial branches can be remarkably improved, the defect of misrecognition or missing division of traditional threshold segmentation is overcome, and the obtained three-dimensional model can be used for aerosol drug deposition simulation, respiratory system disease diagnosis, surgical planning and design and evaluation of personalized inhalation devices.

A fingerprint segmentation method and device is provided. The fingerprint segmentation method may include dividing a fingerprint image into a plurality of sub-images, pooling each sub-image to acquire a feature map of the fingerprint image, and segmenting the feature map based on a region growing algorithm to acquire a fingerprint region in the fingerprint image.

The invention provides a high-precision aphid identification and counting method in a mobile computing scene, and relates to the technical field of agricultural pest intelligent monitoring, and the method comprises the steps: taking a monitored plant as a center, collecting a plurality of groups of visible light images and depth images along an annular path, and constructing a plant three-dimensional point cloud model through coordinate registration and point cloud fusion; segmenting and extracting independent blade units by adopting a self-adaptive region growing algorithm, and carrying out plane fitting and two-dimensional projection to generate a blade emmetronized image; recognizing the envisaged image through a pre-trained aphid recognition model to obtain an aphid bounding box set and a fine-grained category; establishing a mapping relationship between the three-dimensional point cloud and the two-dimensional image, and mapping the bounding box back to the three-dimensional point cloud to obtain an aphid candidate point set; and carrying out double-constrained three-dimensional clustering by utilizing spatial Euclidean distance and fine-grained category label consistency, and screening and marking independent aphid individuals. The method can realize high-precision aphid identification and counting, and can be widely applied to the field of agricultural pest intelligent monitoring.

The invention discloses a heavy mineral placer prediction method, device and equipment based on paleo-coastline reconstruction, and the method comprises the steps: determining the paleo-sea level relative height of a sample region, and generating a target paleo-coastline through a region growing algorithm; generating an ancient landform elevation model, identifying a plurality of ancient landform units based on the topographic position index, and screening out ancient landform exposure units; performing mineral identification on the ancient landform exposure unit to obtain a heavy mineral abundance map; constructing a hydrodynamic-deposition coupling model, identifying a sediment accumulation area of the sample area, and comparing the sediment accumulation area with the heavy mineral abundance map to obtain a heavy mineral enrichment area; dividing a positive sample area and a negative sample area, and training to obtain a heavy mineral placer prediction model; a mineralization probability distribution diagram of the target area is obtained through prediction, and the heavy mineral target area is determined according to the mineralization probability distribution diagram. The method improves the prediction precision of the mineralization probability of the heavy minerals, thereby improving the efficiency and accuracy of positioning and exploration of heavy mineral placers, and can be applied to the technical field of resource detection.

The invention relates to the technical field of medical image processing, and discloses a chronic obstructive pulmonary emphysema distribution quantitative method and system based on multi-scale segmentation, and the method comprises the steps: carrying out the anisotropic diffusion filtering noise reduction of a chest CT image; segmenting a lung field and removing a blood vessel bronchial structure by adopting a region growing algorithm; multi-scale image representation is constructed based on a Gaussian pyramid, an emphysema candidate area is identified in a coarse scale layer, and a boundary is accurately drawn by adopting a self-adaptive threshold value in a fine scale layer; extracting local texture features to distinguish the lobular central emphysema and the total lobular emphysema; dividing severity levels according to spatial aggregation characteristics and density gradient distribution, and calculating an air swelling volume ratio and a distribution heterogeneity index; the three-dimensional pseudo-color volume is used for drawing visualization, a structured quantitative report is generated, accurate segmentation and subtype classification of the emphysema area are achieved, and comprehensive quantitative analysis indexes are provided.

This invention proposes a spatial mapping method and system for new bamboo shoot yield based on satellite-UAV fusion. Addressing the problem of remote sensing observation gaps caused by the underground growth of bamboo shoots, this invention constructs a multi-scale fusion framework of "satellite identification of potential high-yield areas + UAV inversion of old bamboo density." First, a new bamboo yield prediction model is constructed using old bamboo density and topographic factors as variables through ground-based sample plot surveys. Second, indices such as NDVI and LSWI are calculated using time-series imagery, and a decision tree model is used to identify potential new bamboo germination areas in high-yield bamboo forests. Then, based on high-resolution UAV imagery, a region growing algorithm is used to extract individual old bamboo plants and invert the spatial distribution of old bamboo density. Finally, using the potential areas as masks, old bamboo density and topographic factors are substituted to generate a spatial distribution map of new bamboo yield. This invention enables large-area, low-cost, and high-precision mapping of new bamboo shoot yield, providing core technical support for precise bamboo forest management.

The application discloses a kind of wild fire smoke cluster and its boundary vector automatic identification method, system, storage medium and electronic equipment, belong to satellite remote sensing and disaster monitoring technical field.Method includes: obtaining satellite active fire point, fire event and UVAI data;Fire source area is positioned based on fire event geographic location and initial plume pixel is marked;The spatially continuous smoke cluster connected domain is identified by iterative region growing algorithm;According to the specific distribution form of connected domain pixel in satellite pixel matrix, the accurate closed boundary vector is automatically generated using matching geometric algorithm;Finally, using multiple criteria, the spatial isolated independent smoke cluster is screened out.The application avoids the defects that traditional image method is easily disturbed by environment from physical principle, realizes fully automated processing without training data, and can output vector boundary that can be directly used for diffusion and emission evaluation, significantly improves the accuracy, automation degree and application value of wild fire smoke cluster monitoring.

The application relates to an insulator image segmentation method based on an unmanned aerial vehicle infrared enhanced image, and relates to the technical field of image processing, and comprises the following steps: acquiring an infrared image sequence through an infrared enhanced camera to obtain a standard infrared image sequence, performing state recognition by using an image perception prior engine, and outputting a predicted insulator attention heat map and a predicted boundary confidence map; optimizing and adjusting an initial threshold segmentation algorithm and an initial region growing algorithm to obtain an adaptive threshold segmentation algorithm and an adaptive region growing algorithm; performing image segmentation on the standard infrared image sequence to output an insulator image sequence, and performing early warning judgment on the insulator image sequence based on an adaptive early warning mechanism. The application solves the problem that the traditional insulator image segmentation method cannot effectively deal with the problems of low temperature contrast, large noise interference and complex background of the infrared image, so that the insulator segmentation is prone to false segmentation and missed segmentation, and the high-precision requirement of fault detection cannot be met.

The invention relates to the technical field of slope engineering modeling, and discloses a slope structural surface unmanned aerial vehicle point cloud fusion refined modeling method, which comprises the following steps: acquiring slope global point cloud data and optical images through an unmanned aerial vehicle carried high-precision positioning system, combining local point cloud data acquired by a handheld laser scanner, calibrating and pre-processing, and obtaining a slope global point cloud model; a point cloud normal vector is determined based on a local plane fitting technology, structural boundary information is extracted through a normal vector difference boundary detection algorithm, structural plane clustering segmentation is completed by adopting a region growing algorithm, finally, a triangular mesh model is generated, texture mapping and hierarchical simplification are performed, and refined modeling of a slope structural plane is realized. According to the method, the modeling precision and efficiency are improved, and reliable model support is provided for stability analysis and risk assessment of slope engineering.

This invention relates to a method for detecting defects in aircraft skin based on augmented reality (AR) technology. It combines raster projection with a binocular camera to acquire 3D point cloud data of the aircraft surface, and calculates depth information through differential image processing to construct a high-precision 3D point cloud model. Moving least squares (MLS) is used to smooth noise in the point cloud data, and a region growing algorithm is employed to accurately identify defect areas, incorporating local curvature and normal information. Further quantitative analysis is used to precisely calculate the shape, size, and depth of the defects. Augmented reality technology projects defect information onto the aircraft surface in real time, facilitating intuitive location and handling of defects by maintenance personnel. This method requires no coupling agent or radiation source, causing no secondary damage to the aircraft surface. It is simple, fast, efficient, and highly adaptable to various environments. This invention provides a more efficient, accurate, non-destructive, and easy-to-operate defect detection solution with broad application prospects.

The invention provides a ceramic strength intelligent prediction method and system considering micropore spatial configuration. The method comprises the following steps: step 1, acquiring a microscopic image of a potential fracture area on the surface of a structural ceramic sample by using an electron microscope; step 2, carrying out image processing and feature extraction on the microscopic image based on a region growing algorithm to obtain a hole defect set of each sample; step 3, constructing a graph structure based on each sample hole defect set by adopting a method of combining MST, KNN and quantile threshold cutting; and 4, inputting the graph structure into the graph neural network prediction model to obtain the bending strength of the structural ceramic. According to the invention, through explicit coding of micropore spatial configuration and interaction, improvement of accuracy and robustness of strength prediction in a small-sample and non-destructive detection scene is facilitated.

This invention discloses a method and system for defect identification of molybdenum-rhenium alloy pipe fittings based on high-definition microscopic images. The method includes the following steps: acquiring high-definition images of the molybdenum-rhenium alloy pipe fittings using a high-definition microscope; preprocessing the high-definition images of the molybdenum-rhenium alloy pipe fittings to obtain an image of the molybdenum-rhenium alloy pipe fittings to be detected; performing region growing algorithm processing on the image of the molybdenum-rhenium alloy pipe fittings to be detected to obtain a defect region to be detected; identifying defects in the defect region to be detected to obtain the type of defect; and determining that the molybdenum-rhenium alloy pipe fittings has a defect problem based on the type of defect.

This invention discloses a method for extrinsic parameter calibration of a LiDAR and camera based on directional equalization, comprising the following steps: extracting edge feature maps from the camera image and performing Euclidean distance transformation on the edge feature maps to obtain the range field and its gradient; performing voxel downsampling on the LiDAR point cloud and using a region growing algorithm to segment the principal plane from the downsampled LiDAR point cloud; solving for the intersection lines between adjacent principal planes pairwise to obtain LiDAR edge line features; transforming the LiDAR edge line feature points to the camera coordinate system and projecting them onto the image plane, and performing bilinear sampling of the range field and its gradient at the projected pixel positions; constructing a global orientation histogram and calculating the directional equalization weights; constructing a joint objective function that fuses the range field residuals and the directional equalization weights; and iteratively optimizing the joint objective function to solve for the optimal extrinsic parameters between the LiDAR and the camera. The advantage of this invention is that it improves the stability of extrinsic parameter estimation.

The present application relates to the vertical shaft cage detection technical field, specifically, it relates to a kind of industrial intelligent vision detection and diagnosis system for vertical shaft cage, acquisition unit synchronously obtains cage flange surface image and cage car pitch angle, coal dust shadow quantization unit uses bimodal adaptive threshold segmentation combined with region growing algorithm to extract coal dust area, introduce space weight factor to calculate average gray value and coverage area proportion, dynamic coupling calibration unit relies on three-dimensional look-up table and nonlinear interpolation of multi-working condition calibration, match the optimal deviation compensation coefficient of the combination of jitter and coal dust, generate dynamic gray scale segmentation threshold, defect diagnosis decision unit is through zoned adaptive segmentation, multiscale morphological filtering denoising, combined with gradient amplitude and curvature double-constraint subpixel edge detection accurately calculate wear depth, according to real-time depth and historical trend output grading control instruction, eliminate coupling interference, improve cage wear detection accuracy, provide reliable basis for safe operation.

This application discloses a method for detecting corneal thickness after phacoemulsification cataract surgery based on anterior segment OCT, belonging to the field of image analysis technology. The method includes: dividing the cornea into multiple first regions based on a region growing algorithm and initial corneal thickness data; obtaining second regions based on each first region; obtaining multiple partial boundaries based on the second regions; obtaining fibrotic regions based on each partial boundary; and adjusting the scanning density and refractive index of the second regions based on the fibrotic regions and a scanning density correction coefficient to obtain the final corneal thickness data. This application combines the fibrotic regions and the scanning density correction coefficient to adjust the scanning density and refractive index of the second regions, avoiding the errors of calculations using a uniform refractive index in existing technologies, ultimately achieving accurate measurement of corneal thickness and solving the technical problem that existing OCT detection cannot accurately measure corneal regional thickness.

The invention relates to the technical field of infrared detection of power equipment, in particular to a CVT fault detection method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the image enhancement of a collected original CVT infrared image; inputting the enhanced CVT infrared image into a pre-trained CVT infrared image fault classification model to obtain a fault classification result output by the CVT infrared image fault classification model; when the fault classification result is a fault type, performing image segmentation on the CVT infrared image to obtain a region of interest and a background region; performing image segmentation on the region of interest by adopting a region growing algorithm to obtain a fault region and a non-fault region; and calculating a relative temperature difference between the fault area and the non-fault area, and determining a fault level of the fault area based on the relative temperature difference. The technical scheme is used for thermal fault detection of the CVT, unnecessary casualties can be prevented, and the detection cost is reduced.

This invention discloses a non-contact tire deformation recognition method based on a fine-tuned large-scale visual model. The method includes calibrating a monocular high-speed camera, fine-tuning the parameters of the mask decoder of the large-scale visual model using a tire image dataset, generating a point coordinate cue sequence based on pixel calculation using OpenCV, inputting the cue sequence into the fine-tuned large-scale model for segmentation, further processing the pixel matrix using OpenCV to generate a point cue sequence, and post-processing using iterative geometric fitting and region growing algorithms to obtain the mechanical deformation parameters of the target sample tire. This invention can achieve accurate and rapid tire deformation recognition, reaching pixel-level precise segmentation. It solves the problems of large errors, limited measurement environments, nonlinear image distortion caused by cameras, weak generalization ability, and high training costs in existing computer vision-based vehicle tire recognition methods.

The application provides a video multi-level accurate display method and system based on a sight distance segmentation, and relates to the technical field of image processing. The method comprises the following steps: performing adaptive boundary scanning to identify an object boundary, taking one boundary point in the object boundary as a seed point, and cutting out a first heat source area and a second heat source area on a spatial topological structure corresponding to a low light scene through a region growing algorithm driven by a decay component in a thermal gradient matrix; marking depth levels corresponding to the first heat source area and the second heat source area according to intensity mutation characteristics of a mutation component and exponential decay characteristics of the decay component in the thermal gradient matrix; performing corresponding adjustment on the first heat source area and the second heat source area respectively based on the depth level marking, and fusing the adjusted first heat source area and the second heat source area to generate an optimized display image. The application can realize multi-level accurate display of a video in a low light environment.

The invention discloses an intelligent identification method for hot spots of a photovoltaic module. The method comprises the following steps: acquiring multi-scene original image data of a to-be-identified photovoltaic module; preprocessing the original image data to obtain a standardized image set; constructing a multi-scale image registration model based on a deep convolutional neural network, and realizing pixel-level alignment of the to-be-registered image and the reference image; and carrying out hot spot candidate region preliminary extraction on the registered image set by adopting a region growing algorithm. The invention relates to the technical field of photovoltaic module hot spot intelligent identification. According to the intelligent identification method for the hot spots of the photovoltaic module, pixel-level alignment of the image to be registered and the reference image is realized through the multi-scale image registration model constructed by the deep convolutional neural network, the hot spot area in the photovoltaic module is accurately identified, the problems of misjudgment and missed judgment possibly existing in a traditional method are avoided, and the identification accuracy of the hot spots in the photovoltaic module is improved. And long-term stable operation and performance optimization of the photovoltaic power station are facilitated.

The invention discloses a cataract ultrasonic emulsification postoperative cornea thickness detection method based on anterior segment OCT, and relates to the technical field of image analysis. The method comprises the following steps: dividing a cornea into a plurality of first regions based on a region growing algorithm and initial cornea thickness data; acquiring a second area based on each first area; acquiring a plurality of partial boundaries based on the second region; based on the boundaries of all the parts, a fibrosis area is obtained; and adjusting the scanning density and the refractive index of the second region based on the fibrosis region and the scanning density correction coefficient to finally obtain cornea thickness data. According to the method, the scanning density and the refractive index of the second area are adjusted by combining the fibrosis area and the scanning density correction coefficient, so that the error of calculation by using a unified refractive index in the prior art is avoided, accurate measurement of the cornea thickness is finally realized, and the technical problem that the thickness of the cornea area cannot be accurately measured through existing OCT detection is solved.