123 results about "Viewing frustum" patented technology

The invention discloses a scene rendering method and system based on three-dimensional Gaussian splashing. The method comprises the following steps: organizing and constructing an original 3D Gaussian set to obtain a spatial hierarchical structure; generating corresponding level details for the primitives in the spatial hierarchical structure to obtain a spatial hierarchical structure associated with the level details; traversing the spatial hierarchical structure associated with level details, and executing hierarchical view cone cutting and shielding elimination to obtain a visible node list; traversing the visible node list, calculating a level detail selection standard, and generating a level detail activity Gaussian set; performing optimization sorting on the level detail activity Gaussian set to obtain an activity primitive list; and performing tile-based rasterization on the active primitive list, and performing adaptive processing according to level details of the primitives to obtain a final color value of each pixel. According to the method, the rendering performance can be greatly improved, the occupation of a memory and a video memory is remarkably reduced, the rendering quality is improved, visual flaws are reduced, and the expandability of the 3DGS rendering method is enhanced.

Disclosed in the present invention is a three-dimensional Gaussian splatting optimization method for an unposed input. The method comprises: for an input image, predicting a ray bundle distribution by using a ray prediction model, to obtain distribution features in the form of ray bundles; on the basis of the distribution features in the form of ray bundles, calculating a camera pose; for the ray bundle distribution, performing sampling on the basis of the volumetric density of light rays, to obtain an initial spatial distribution of a three-dimensional Gaussian point cloud focused on a visual center area; for the input image, obtaining a visible shell by means of view frustum projection and object-mask computation; on the basis of the initial spatial distribution of the three-dimensional Gaussian point cloud and the visible shell, performing three-dimensional Gaussian splatting scene training, to obtain a three-dimensional scene reconstruction model satisfying a preset loss function criterion, the loss function comprising a training regularization term for a camera pose parameter. The present invention provides important scene initialization information for three-dimensional Gaussian splatting training, and significantly improves the quality and the richness of detail of the final three-dimensional structure.

The invention relates to the field of scene rendering, and particularly discloses a three-dimensional scene optimization method and system for collaborative rendering of dynamic LOD and view cone rejection based on space-time prediction, and the method comprises the steps: dynamically calculating the visibility frequency of an object, and dynamically adjusting the loading and rendering modes of objects with different priorities; predicting a view cone range of multiple frames in the future by using an LSTM network architecture; dividing the scene into uniform grid blocks, and then performing grading elimination; optimizing a heterogeneous computing pipeline; rendering the execution process; the system comprises a dynamic LOD and visual cone rejection collaborative optimization module, an LSTM visual cone prediction module, a block-level mixed rejection module, a heterogeneous calculation pipeline module, an optimization CPU-GPU task allocation and data transmission module and a rendering execution control module. According to the method, the technologies of collaborative optimization of dynamic LOD and view cone removal, block-level mixed removal, heterogeneous calculation assembly line optimization and the like are adopted, so that unnecessary rendering calculation and resource loading are reduced, and the rendering efficiency is improved.

The invention relates to an OpenGL (Open Graphics Library)-based large-scale three-dimensional model object quick pickup method. The method comprises the following steps: converting an interactive operation of a user into a three-dimensional space structure; a CPU constructs octree space acceleration structures based on a bounding box of an object in a large-scale three-dimensional model, and leaf nodes, intersecting with a pickup ray or a pickup view pyramid, of an axis alignment bounding box in each octree space acceleration structure are screened out to serve as candidate leaf nodes; collecting triangles contained in all candidate leaf nodes, and constructing a candidate triangle set; dynamically allocating thread groups and threads of the GPU based on the number of candidate triangles in the candidate triangle set; and performing highlight display on the candidate triangles intersected with the pickup ray or the pickup view cone, and feeding back IDs of the corresponding candidate triangles. According to the method, the data transmission quantity and the video memory access pressure are remarkably reduced, the pickup response time is greatly shortened in a large-scale three-dimensional model scene, and the requirements of real-time interaction and visual operation can be met.

The invention discloses a rendering method, which comprises the following steps: constructing an octree, and storing a to-be-rendered object in a to-be-rendered three-dimensional scene in a child node of the octree to obtain a space management octree; dividing the view cone into different view cone levels; traversing the space management octree, and determining a first intersection part between the space of the child node and the view cone and a view cone level to which the first intersection part belongs; distributing the first intersection part to different rendering queues according to the view cone hierarchy to which the first intersection part belongs; shielding and removing the first intersection parts in the different rendering queues to obtain current visible objects to be rendered; and rendering the current visible to-be-rendered object. According to the method, the octree is constructed to manage the three-dimensional scene space, the view cone levels are divided, the rendering queues are distributed, the view cone removing and multi-level shielding removing technologies are combined, the visible objects are accurately judged, the rendering load is effectively reduced, and the method has the advantages that the rendering performance is improved, space management is optimized, and the complex scene processing efficiency is improved.

The embodiment of the invention provides an underwater scene three-dimensional reconstruction method based on three-dimensional Gaussian sputtering and related equipment, and belongs to the technical field of computer vision. The method comprises the following steps: performing feature point analysis on each real two-dimensional image to obtain a sparse point cloud set; a three-dimensional Gaussian ellipsoid of each point cloud in the sparse point cloud set is initialized to represent a three-dimensional underwater scene, and the three-dimensional Gaussian ellipsoid is used for representing Gaussian distribution of the position, the covariance matrix, the color and the opacity of the point cloud; selecting a plurality of three-dimensional Gaussian ellipsoids located in a corresponding camera view cone based on the target camera pose, and performing two-dimensional projection rendering according to the plurality of selected three-dimensional Gaussian ellipsoids to obtain a rendered two-dimensional image; and updating the three-dimensional Gaussian ellipsoids in the point cloud space according to the rendered two-dimensional image and the real two-dimensional image with the same target camera pose to obtain a plurality of three-dimensional Gaussian ellipsoids accurately representing the underwater scene. According to the invention, the accuracy and efficiency of underwater scene three-dimensional reconstruction can be improved.

The invention discloses a port avionics hub digital twin multi-level simulation model construction and visualization method, which comprises the following steps: obtaining a multi-view RGB image of a target port area, obtaining a camera attitude of each view by using an SfM algorithm, and reconstructing a sparse three-dimensional point cloud; constructing an LOD2 shell model with a semantic mark; an LLO3 model with holes is constructed; the LOD0-LOD3 model is serialized into a glTF 2.0 binary format and transmitted to a client side through HTTP / 2, the client side calls WebGL 2.0 through a browser to complete decompression, uploading and PBR rendering, and interactive plug-in-free visualization is achieved; constructing an eight-fork-two-fork-four-fork mixed spatial index, and performing multi-stage division and persistence on the LOD < 0 >-LOD < 3 > tiles, so that a client locates visible tiles within constant time based on node numbers; visual cone cutting and screen space error judgment are adopted to execute progressive loading on visible tiles, LOD0 is loaded at first, then LOD1-LOD3 is subjected to incremental superposition, low-level tiles in the same area are replaced with the LOD3 in a cross fading mode at the ultra-close distance (the Euclidean distance d from the center of a camera to the center of a target tile is smaller than or equal to 100 m), and smooth visual experience is obtained in a browser.

The invention provides a water conservancy twinborn model rendering optimization method for dynamically generating gray level image frames based on a view cone. The method comprises the following steps: obtaining visible area model data after the current view cone is dynamically cut; outputting the visible area model data to a multi-layer rendering target comprising a color buffer area, a vertex buffer area and a depth buffer area through off-screen rendering so as to generate intermediate frame data isolated from screen display; performing gridding processing on the intermediate frame data, and displaying the equipment resolution according to the physical range size of the water conservancy project or the drainage basin scene; the average color value of each grid unit is calculated, desaturation processing is carried out to generate a basic grey-scale map, and adaptive feature enhancement processing is carried out to generate an optimized grey-scale map or a gradient map; and a GPU rendering pipeline is input and acts on hydrological feature analysis of the post-processing stage of the current frame and a preloading data source rendered by a twinborn model of the next frame at the same time, historical frame data are reused in a static region through motion detection, and real-time rendering is triggered in a dynamic region.

The invention discloses an ultrahigh-resolution immersive scene rendering method based on hexagonal picture segmentation. The method comprises the following steps: 1, determining a hexagonal segmentation style of an immersive display surface; 2, unfolding the immersive display surface into a plane; 3, acquiring 3D point information of each hexagonal surface, respectively calculating the width W and the height H of the enveloping quadrilateral surface, and calculating a local coordinate system V of the hexagonal surface S; 4, using W, H and V to construct a view cone and a viewpoint matrix of the three-dimensional scene; 5, setting a shade M of a plane expansion view for the hexagonal surface S; 6, calling a rendering scene of the 3D scene for the mask region by using information of the view cone and the viewpoint matrix, and rendering the mask region into a picture at the position S of the hexagonal surface; and 7, repeating the steps 5-6, and sequentially carrying out the same treatment on each hexagonal surface. According to the method and the device, the ultrahigh-resolution picture can be rendered in a scene segmentation form, and the method and the device are suitable for ultrahigh-resolution picture rendering under immersive display equipment.

The invention discloses a roadside multi-mode fusion sensing method. According to the method, an initial 3D view cone tensor with depth information is generated through a depth estimation network, features of pixel points are projected to a coordinate system in combination with internal and external parameter matrixes of a camera, then a three-dimensional space point coordinate set is generated according to projected feature position information, a Gaussian function is adopted to distribute weights of neighborhood grids, and a three-dimensional space point coordinate set is obtained. And finally, inputting the image features and the features into a fusion unit, respectively carrying out uncertainty modeling on the two types of features, automatically generating a fusion weight, and finally, carrying out normalization weighted aggregation on the features in the neighborhood so as to reduce the position approximation error and obtain the calibrated image features, and finally, inputting the image features and the features into the fusion unit, and respectively carrying out uncertainty modeling on the two types of features and automatically generating a fusion weight. And dynamic optimization is carried out according to a self-supervised loss function in the fusion process, so that the stability of multi-modal feature fusion is ensured, the detection accuracy is high, the performance is relatively good, and the problem of global feature dislocation caused by external parameter deviation and depth error can be avoided.

The invention discloses a valve recognition pose estimation method and system based on an FPPE network, and relates to the field of image recognition. The problems that an existing visual positioning and pose detection method for the valve is not high in pose recognition degree and too large in calculated amount are solved. The method comprises the following steps: collecting a multi-angle valve picture and valve point cloud data, and constructing a valve point cloud data set and template point cloud; the RT-DETR v2 target detection network is trained; an FPPE network is constructed and trained; according to the trained RT-DETR v2 network, carrying out target identification and plane target position detection on the valve image to obtain a central position coordinate of a valve area and an ROI boundary point pixel coordinate; points of all depths in the ROI space area are extracted to form a view cone point cloud; and inputting the view cone point cloud into the trained FPPE network to obtain a final valve pose estimation result. The method is applied to the technical field of object positioning.

The invention discloses a holographic computing task distributed distribution method and system based on multi-person multi-view interaction, and relates to the technical field of distributed computing and holographic interaction.The system comprises a main scheduling node, a plurality of computing nodes and a user terminal.The main scheduling node obtains scene graph data of a current frame and view cone parameters of the user terminal; generating a dynamic granularity task block with metadata through a dynamic sensing task block generator; obtaining a visual angle coherence cluster through a visual angle coherence clustering engine; calculating sharing benefits and full-link cost of task blocks and clusters, constructing a benefit-cost ratio objective function, and obtaining an optimal matching relationship through an optimization algorithm; according to the method, dynamic adaptation of task granularity and optimal allocation of resources are achieved, redundant rendering and transmission delay are reduced, and the fluency of multi-user holographic interaction and the resource utilization rate are improved.

This invention discloses a method, apparatus, device, and storage medium for constructing a 3D map, applicable to the field of 3D mapping. It involves determining a reference point within the view frustum of a viewpoint, constructing a reference plane containing the reference point and parallel to the view frustum, determining the height of the 3D map rendering interface window, and determining the height of the reference plane window. The unit of the 3D map rendering interface window height is pixels, while the unit of the reference plane window height is length. A distance-pixel ratio is determined based on the ratio of the reference plane window height to the 3D map rendering interface window height. The 3D map scale of the reference point is determined based on the distance-pixel ratio, constructing a 3D map containing the 3D map scale. Based on perspective projection rules, the scale information at any location in the 3D map can be calculated quickly and accurately, improving the efficiency of constructing a 3D map with a high-precision scale. The high-precision scale can assist in obtaining accurate geographic information within the 3D map, thereby improving modeling accuracy.

The application discloses a virtual camera simulation data generation method and system based on a Gaussian point cloud model, relates to the technical field of three-dimensional point cloud processing and modeling, and comprises the following steps: after completing multi-source data interaction, a point cloud dataset is established according to the data interaction result; after completing space-time registration on the dataset, a fusion space dataset is constructed, and a Gaussian point cloud model is obtained through modeling; then, virtual camera parameters are configured, a frustum is constructed to identify invalid points, and point labels are established; center perception fusion is carried out by using the point labels to obtain a perspective simulation image; and after target perception feedback correction, simulation data is established. The application solves the technical problem that the prior art cannot accurately fuse multi-source information, the virtual camera simulation data generated has a large deviation from a real scene, and the data processing result is inaccurate, and achieves the technical effect that the virtual camera simulation data accurately fits the real scene and the accuracy of the data processing result is improved.

This invention discloses a 3D train detection and positioning method based on ground-based monocular vision. It eliminates the need for satellite signals or additional trackside equipment. Through convolutional-trans depth estimation and pseudo-point cloud optimization, the absolute relative error of depth estimation can be reduced to below 0.06. After pseudo-point cloud optimization, the 3D detection accuracy is improved by more than 20%, achieving a final positioning accuracy of 1 meter, meeting the auxiliary positioning requirements of train control systems. Furthermore, through view frustum region extraction and lightweight network design, the inference frame rate is ≥10fps, meeting the real-time positioning requirements of high-speed trains. In addition, existing ground-based monocular cameras can be reused, eliminating the need for additional hardware. The deployment and maintenance costs are only 1 / 10 of those of transponder positioning systems. Overall, this invention provides a novel technical path for train positioning and can be widely applied to auxiliary positioning systems for conventional and high-speed railways. It is particularly suitable for the intelligent upgrading and transformation of existing lines, possessing extremely high engineering application value.

The invention relates to the technical field of computer graphics and geological exploration visualization, and discloses a geologic body space penetration processing system based on a domestic digital twin engine technology, and the system comprises a heterogeneous data mapping module which is used for carrying out the normalization interpolation of discrete geological data, and generating a continuous three-dimensional body texture; the view cone depth weighted statistical analysis module is used for carrying out view cone sampling on the body texture and generating a weighted histogram in combination with a depth weight; the time domain damping adaptive threshold value control module is used for solving an instantaneous threshold value based on the histogram and generating a smooth final application threshold value in combination with a historical threshold value; and the multi-dimensional gradient mapping and penetration rendering module is used for calculating comprehensive visual potential energy in combination with the attribute value and the surface orientation factor, and executing fragment elimination or output based on a final threshold value. According to the method, a time domain damping self-adaptive threshold control mechanism is constructed, the time sequence stability of a rendered picture in the dynamic roaming process is ensured, and visual jump is eliminated.

This application discloses a method, apparatus, and electronic device for displaying a three-dimensional model of a game character. The method includes: determining a target three-dimensional model and multiple preset original texture layers; obtaining a resizing parameter for the two-dimensional original texture layer within the viewing frustum based on the relative position of the two-dimensional original texture layer and a virtual camera; adjusting the original size of the two-dimensional original texture layer to a target size based on the resizing parameter to obtain a target texture layer; generating a three-dimensional virtual scene background based on the target texture layer, and displaying the target three-dimensional model within the three-dimensional virtual scene background. This application solves the technical problem of virtual scene backgrounds lacking a sense of three-dimensionality.

The invention discloses a management-oriented data center system and a data processing method. The system comprises a 3D visualization module, a resource monitoring module, a machine room management module, an interaction adaptation module, a feedback closed loop module, a system configuration module and a data processing unit. According to the method, the physical layout and the operation state of the data center are visually presented through the 3D visualization technology, and the problem that traditional management information is not visual is solved; the refined resource monitoring and machine room management functions improve the resource allocation and scheduling efficiency; the responsive interactive design and shortcut key support optimize the operation convenience; the feedback closed loop module realizes quick response of user requirements and continuous optimization of the system; a system configuration module and a security mechanism guarantee the stability, expandability and data security of enterprise-level applications, and in addition, through performance optimization strategies such as view cone elimination, assembly lazy loading and virtual rolling, it is ensured that the system can still run smoothly in a large-data-volume scene.

The application discloses a kind of based on three-dimensional reconstruction's silt flocculation structure parameter measurement method, system, equipment and medium, belong to flocculation monitoring technical field.The method comprises: by at least two image acquisition devices, the two-dimensional image of silt flocculation is synchronously collected from different visual angle, and its imaging geometric parameter is obtained;The two-dimensional image of silt flocculation is segmented to obtain the binary mask of each visual angle;Based on imaging geometric parameter, each mask is reversely projected to three-dimensional space to form visual cone, and the three-dimensional visual convex hull model of silt flocculation is obtained by intersection operation;Based on the three-dimensional visual convex hull model, the structure parameter of silt flocculation is output.Therefore, by implementing the present application, the problem that the calculation error of irregular silt flocculation structure parameter is large in the prior art based on single visual angle two-dimensional projection and sphere assumption can be solved, and the accurate measurement of the real three-dimensional morphology of silt flocculation is realized.

The application provides a virtual reality building roaming method and system based on scene rendering, relates to the technical field of virtual reality, and comprises the following steps: dividing building three-dimensional model data by means of an octree space segmentation algorithm, and determining a detail level threshold according to a viewpoint distance to construct a multi-level detail model. A basic light map is baked by means of a pre-computed irradiance transfer technology, and is combined with a multi-resolution shadow map to perform real-time synthesis. A physical-based rendering method is adopted to calculate light, and dynamic light rendering data is generated by fusing the light and shadow synthesis data of a scene. A current viewpoint is determined based on spatial position and posture data of a virtual reality head-mounted display, view frustum culling and occlusion culling are performed, and GPU instance rendering is performed on visible building data. A rendering image is optimized by means of high dynamic range processing and temporal anti-aliasing technology, and a final scene image is generated by combining an atmospheric scattering algorithm. The application can effectively improve the rendering efficiency and scene realism of virtual reality building roaming.

The invention relates to a high-precision terrain generation method, device and equipment based on dynamic paging. The method comprises the following steps: dividing global terrain data into multi-level tiles, generating multi-level LOD data, and constructing a quadtree-hash table two-level index structure; based on an index structure, dynamically calculating an LOD level according to a camera view cone and a predicted viewpoint position, and generating a paging scheduling queue according to a priority weight model; asynchronously loading high-priority paging data in an independent loading thread, and performing cache management by adopting an LRU strategy; and carrying out real-time three-dimensional rendering on the loaded pages, realizing smooth switching by adopting interpolation transition, carrying out vertex weighted average and texture mixing processing on adjacent page boundaries, and carrying out cyclic mapping on longitude boundaries. According to the scheme, the problems of high memory occupation, low rendering efficiency, loading delay and the like in large-range high-precision terrain real-time generation are effectively solved, and the system performance and the user experience are improved.

This invention discloses a deep learning-based 3D reconstruction method for ship engines, belonging to the field of computer vision technology. The method includes: obtaining 10 input images containing a reference image and neighboring images from 10 images; inputting each of the 10 input images into a pre-trained DAS-MVSNet network, which performs the following steps: sequentially inputting each input image into a feature extraction module, a Deep-ASPP module, and a Conv1 convolutional layer to obtain 10 first feature maps; establishing a reference camera viewpoint and uniformly sampling along the optical center direction within its view frustum to form multiple planes of different depths; mapping each first feature map to multiple planes of different depths to generate 1 feature volume; aggregating the 10 feature volumes into a single cost volume and inputting it into a four-scale 3D convolutional neural network to obtain an optimized depth map; and finally, after traversing all images, generating a point cloud model based on the optimized depth map. This invention improves the accuracy and completeness of 3D reconstruction.

The invention discloses a point cloud map rendering method based on a SceneKit framework. The method comprises the following steps: performing structured processing on local point cloud data, and determining preprocessed local point cloud data; based on the preprocessed local point cloud data, a first vertex source object and a rendering mode type under a SceneKit framework are determined, the first vertex source object is associated with a data buffer area, and the rendering mode type is a point cloud rendering type; on the basis of a camera of a SceneKit frame, region division is carried out on the first vertex source object by utilizing a view cone cutting algorithm, and a second vertex source object of a to-be-rendered region is determined; generating single drawing call data based on buffer data corresponding to the second vertex source object in the data buffer area; and generating a rendered image based on the single drawing call data and the rendering mode type. According to the method, structured preprocessing of the point cloud data, point primitive geometry construction and a built-in performance optimization mechanism are realized through the SceneKit framework, the development threshold is reduced, the rendering efficiency is improved, and the memory stability is ensured.

The application discloses a cloud rendering message communication interaction method for connecting multi-end digital twin business scenarios. The method is realized based on a three-layer architecture system of a client layer, a middleware layer and a server layer. The client layer collects the parameters of the three-dimensional camera's view cone in real time; the middleware layer uses the parameters, combines spatial semantic index and distance attenuation law, calculates the dynamic scheduling weight of massive concurrent messages, and accurately distributes them to high-optimal transmission, smooth flow limiting or aggregation degradation channels; at the same time, the server layer reuses the same set of view cone parameters to implement cascade rendering degradation and physical simulation control on twin objects outside the field of view and in the buffer zone. The application not only effectively eliminates the head-of-line blocking effect at the network transmission level, but also realizes the spatio-temporal coupling linkage of message scheduling and computing power distribution based on unified spatial perception, and supports seamless switching of heterogeneous cloud rendering platforms.

This invention discloses a method, system, and terminal device for displaying a 3D geographic information system based on multiple screens, relating to the field of 3D display technology. The method includes: acquiring 3D geographic information data; constructing a 3D geographic scene in a 3D rendering engine; determining the observer's viewpoint within a physical space, which is enclosed by multiple planar display screens forming a multi-screen display; deploying a corresponding number of virtual cameras facing the corresponding screens, with the observer's viewpoint as the origin; configuring the cameras in a physical trapezoidal combination based on the physical size and position of the screens relative to the observer and calculating the asymmetric view frustum projection matrix; allocating rendering tasks to pre-built distributed rendering cluster nodes, and controlling the synchronous output of each camera's image to the corresponding screen through a hardware synchronization mechanism. This invention achieves naked-eye stereoscopic display on multiple screens, eliminates perspective distortion at screen junctions, ensures seamless synchronization of multi-screen images without tearing, and enhances the immersiveness and accuracy of 3D geographic information display.

This application discloses a lightweight preprocessing method for geometric components in 3D scene rendering, relating to the field of 3D rendering. The method includes: when there are multiple edge points in a triangular mesh model, performing component edge shape detection based on the position information of the multiple edge points to determine whether the component to be processed is a cylindrical component; when the component to be processed is determined to be a cylindrical component, performing circular fullness detection on the bottom surface to determine whether the cylindrical component is a cylindrical pipeline; converting the geometric features of each cylindrical pipeline into an instantiated reuse matrix; combining the view frustum parameters of the current viewpoint and the instantiated reuse matrix to perform view frustum culling and occlusion judgment to determine whether each cylindrical pipeline is visible; for cylindrical pipelines, collaborative optimization is achieved in three stages: recognition efficiency, lightweight transmission, and rendering visibility screening. Through the preprocessing mechanism in the early stage of rendering, the overall efficiency of subsequent rendering is effectively improved.

The application discloses a three-dimensional target detection method based on RGBD data and a view cone, and comprises the following steps: acquiring a feature image of a scene RGB image according to the scene RGB image; inputting the feature image of the scene RGB image into a region generation network to acquire a bounding box of a to-be-detected object and a category vector of the to-be-detected object; acquiring a 3D view cone point cloud according to the bounding box of the to-be-detected object and a depth image of the scene; inputting the position coordinates of points in the 3D view cone point cloud and the category vector of the to-be-detected object into a PointNet network with an attention mechanism to acquire the center point coordinates of the to-be-detected object; and further acquiring the parameters of a 3D bounding box of the to-be-detected object to detect the three-dimensional to-be-detected object. The application simultaneously considers the plane information and the depth information of a scene, improves the detection precision, and makes the detection result more accurate while ensuring the efficiency.

The invention relates to the field of image processing, in particular to an industrial gas leakage detection method based on infrared thermal imaging, which comprises the following steps of: acquiring video streams by using a plurality of infrared thermal imagers, and constructing a gas basic geometric voxel grid through a background difference and view cone intersection method; thirdly, the sight line penetration depth and a statistical variable coefficient are calculated, and adaptive normalized radiation feature mapping based on statistical moments is constructed so as to solve errors caused by traditional homogeneous hypothesis; then, measuring and calculating a three-dimensional form compactness factor, generating a heterogeneous space density field weight in combination with the relative depth, and fitting a real diffusion gradient; and finally, calibrating a physical mapping coefficient by using single-point sensor data, and performing integral calculation on the real-time mass flow in combination with a three-dimensional optical flow method. According to the invention, the complex density gradient in the gas can be effectively reduced, and high-precision three-dimensional visual display and quantitative flow monitoring of gas leakage are realized.