3D Model Creation from Sparse Point Clouds Using Image Superposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing three-dimensional modeling technologies using fixed laser scanners face challenges in creating accurate models of objects with unevenly spaced point clouds and small diameters, such as cables on utility poles, due to sparse point cloud data at longer distances.
Innovation Solution
A method and apparatus that superimpose three-dimensional models on images, select relevant point cloud data, and interpolate points to create a complete model, allowing for the creation of three-dimensional models even with unevenly spaced inter-point distances and partial point clouds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed three-dimensional laser scanner is used to acquire point cloud data, then the measurement can be performed from a stationary position, but the point cloud becomes sparse at long distances making it difficult to create 3D models of small diameter objects
Solution Approach 1:
The patent combines point cloud data from a fixed laser scanner with image data from a camera to create a complementary dataset. The image provides visual information that supplements the sparse point cloud data, enabling accurate 3D modeling of small diameter objects even when point cloud density is insufficient alone.
Solution Approach 2:
The patent introduces an image as an intermediary element that mediates between the sparse point cloud data and the target object. By superimposing the initial 3D model with the image and extracting additional point cloud data from regions where they do not overlap, the image serves as a bridge to recover missing geometric information.
2Ease of manufacture
If points are interpolated until the distance between point clouds reaches a certain threshold, then scan lines can be formed, but no points can be interpolated when the distance is large and points are not regarded as being on the same target object
Solution Approach 1:
The patent performs preliminary 3D model creation using available point cloud data before attempting to recover missing data. This initial model serves as a reference framework that guides subsequent image-based data extraction and model refinement, enabling progressive improvement rather than attempting complete reconstruction from scratch.
Solution Approach 2:
The patent implements a feedback mechanism where the initially created 3D model is superimposed with the original image, and the discrepancy between them is used to identify and extract additional point cloud data. This feedback loop continues until the model accurately represents the target object, allowing recovery of small diameter features that were initially missed.
3Quantity of substance
If only partial point cloud data is available, then the data processing load is reduced, but the three-dimensional model cannot be completed accurately
Solution Approach 1:
The patent transitions from relying solely on three-dimensional point cloud data to incorporating two-dimensional image data as an additional dimension of information. This multi-dimensional approach allows the system to recover missing 3D geometric details by leveraging visual information from the image, achieving complete and accurate 3D models even when point cloud data is partial.
Data Source
AI summary
An object of the present disclosure is to enable a three-dimensional model to be created even for a target object which has unevenly spaced inter-point distances and only a part of a point cloud. According to the present disclosure, there are provided an apparatus and a method in which a three-dimensional model of a target object is created from point cloud data in which each point represents three-dimensional coordinates, the three-dimensional model is superimposed on an image in which a target object of the three-dimensional model is photographed, a point cloud data to be added to a point cloud data that constitutes the three-dimensional model is selected by comparing the three-dimensional model with the target object in the image, and the three-dimensional model of the target object is created again using the point cloud data including the point cloud to be added.