Voxel Filtering for Materials Detection Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Explosive detection systems experience latency due to the large size and complexity of multi-dimensional digital images, which slow down data transfer and analysis, particularly when dealing with images containing non-target materials like air and air-like substances.
Innovation Solution
The system generates a second multi-dimensional digital image that disregards voxels representing non-target materials, compressing the image by removing or replacing values associated with air and low-density materials, allowing for faster transfer and analysis of target materials.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the complete multi-dimensional digital image is transferred and analyzed, then detection accuracy is maintained, but system latency increases due to large data volume
Solution Approach 1:
The patent extracts and removes voxels representing non-target materials (air and air-like substances) from the multi-dimensional digital image before transfer and analysis. By taking out only the relevant high-density material voxels that could contain target materials, the data volume is significantly reduced while maintaining detection accuracy for explosives and hazardous materials.
Solution Approach 2:
The patent applies local quality by differentiating between different regions of the image based on material density. Instead of uniformly processing all voxels, the system selectively processes only those voxels with high-density characteristics that are likely to contain target materials, while discarding low-density air voxels that contribute to data volume but not to detection accuracy.
2Reliability
If all voxels including air and low-density materials are processed, then complete image analysis is performed, but data transfer time and processing load increase
Solution Approach 1:
The system extracts and removes voxels representing non-target materials (air and air-like substances) from the multi-dimensional digital image before transfer and analysis. By taking out only the relevant high-density material voxels that could contain target materials, the data volume is significantly reduced while maintaining detection accuracy for explosives and hazardous materials.
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of the image data - specifically, only the high-density voxels that are likely to contain target materials. This partial processing approach maintains reliability for detecting explosives and hazardous materials while significantly improving throughput by avoiding unnecessary processing of air voxels.
3Quantity of substance
If the full-resolution multi-dimensional image is compressed using traditional methods, then data size is reduced, but lossless compression increases processing complexity
Solution Approach 1:
The patent extracts and removes voxels representing non-target materials (air and air-like substances) from the multi-dimensional digital image before transfer and analysis. By taking out only the relevant high-density material voxels that could contain target materials, the data volume is significantly reduced while maintaining detection accuracy for explosives and hazardous materials.
Solution Approach 2:
The patent applies parameter changes by filtering voxels based on their density values. The system changes the data representation from including all voxel types to including only high-density voxels that are likely to contain target materials. This parameter-based filtering achieves data size reduction with simpler processing compared to traditional lossless compression methods.
Data Source
AI summary
A first multi-dimensional digital image of a scan region is generated. The scan region is included in a materials-detection apparatus and is configured to receive and move containers through the materials-detection apparatus. A pre-defined background range of values is accessed, the background range of values representing a range of values associated with non-target materials and the background range of values being distinct from values associated with the target materials. A value of a voxel included in the multi-dimensional digital image is compared to the background range of values to determine whether the value of the voxel is within the background range of values. If the value of the voxel is within the background range of values, the voxel is identified as a voxel representing a low-density material. A second multi-dimensional digital image that disregards the identified voxel is generated to compress the first multi-dimensional digital image.