Bounding Box Pixel Color Analysis for CAD Visibility Culling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CAD systems face performance issues due to the resource-intensive process of loading and visualizing complex 3D models, particularly in cloud-based SaaS implementations, where determining visible and occluded CAD parts is time-consuming and demanding for resource-limited thin clients.
Innovation Solution
The implementation of bounding box-based visualization technology, which assigns unique pixel colors to CAD part bounding boxes, allowing for efficient occlusion culling and visible part determination through pixel color analysis of rendered 2D images, reducing the need for full-detail geometry processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full-detail 3D geometry processing is used to determine visible and occluded CAD parts, then measurement precision is improved, but computing resource consumption increases and processing speed decreases
Solution Approach 1:
The patent segments the complex 3D geometry processing task into two distinct phases: (1) bounding box analysis phase where simplified rectangular envelopes are used for rapid visibility culling, and (2) detailed geometry rendering phase where full precision is applied only to visible parts. This segmentation allows the system to achieve both fast initial filtering and accurate final rendering.
Solution Approach 2:
The patent applies partial action by performing complete precision processing only on the subset of CAD parts that are determined to be visible, while using approximate bounding box analysis for the initial screening of all parts. This avoids the excessive computation that would result from applying full precision to all parts regardless of visibility.
2Measurement precision
If full-detail 3D geometry data is loaded and processed, then visualization accuracy is improved, but network latency and data transmission time increase
Solution Approach 1:
The patent extracts and transmits only the essential bounding box parameters (minimum and maximum coordinates) of CAD parts over the network, rather than transmitting complete 3D geometry data. The client system then performs visibility determination locally using these compact representations, and only requests full geometry data for parts that are determined to be visible, thereby minimizing network transmission time.
Solution Approach 2:
The patent performs preliminary visibility analysis using bounding boxes before initiating the transfer of detailed geometry data. This preliminary action filters out occluded parts in advance, so that network bandwidth is only consumed by transmitting data for visible parts, significantly reducing overall network latency.
3Measurement precision
If complex geometrical and spatial analyses are performed on 3D geometry, then occlusion determination accuracy is improved, but device resource consumption increases
Solution Approach 1:
The patent creates simplified copies of the original 3D geometry in the form of axis-aligned bounding boxes that enclose each CAD part. These bounding box copies retain the essential spatial information needed for occlusion determination while requiring minimal computational resources to manipulate. The system performs all complex spatial analyses on these simple rectangular representations rather than on the original complex geometries.
Data Source
AI summary
A client computing system may include a model visualization engine configured to visualize a view of a computer-aided design (CAD) model in an application window, including by accessing bounding box data for non-visualized CAD parts of the CAD model, assigning a color value to each of the bounding boxes of the non-visualized CAD parts, capturing a 2D image of the view of the CAD model rendered using the bounding boxes for the non-visualized CAD parts, and analyzing the 2D image to identify pixel colors present in the 2D image to determine visible CAD parts in the view of the CAD model. The model visualization engine may further visualize the view of the CAD model by retrieving visualization data for the visible CAD parts and visualizing, in the application window, the visible CAD parts in the view of the CAD model via the retrieved visualization data.