Object Correlation Across Disparate Fields of View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital image processing systems fail to effectively detect and track objects across disparate fields of view, particularly when images are captured by different recording devices with non-overlapping or partially overlapping fields of view, leading to challenges in object identification and correlation across varying perspectives.
Innovation Solution
A data processing system equipped with an object detection component, an object classification component, and an object matching component uses locality sensitive hashing and an inverted index central data structure to detect objects, classify them hierarchically, generate descriptors, and correlate objects across different images from various recording devices, even if they have different fields of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If objects are detected and tracked across disparate fields of view using traditional methods, then object identification is attempted, but correlation accuracy deteriorates due to non-overlapping perspectives and varying recording device viewpoints
Solution Approach 1:
The system transforms objects into parameter-based descriptors by extracting hierarchical classification categories (e.g., object type, color, shape, size) that remain consistent across different field of views. These parameterized descriptors enable accurate correlation by comparing intrinsic object properties rather than relying on perspective-dependent visual features.
Solution Approach 2:
The patent introduces an intermediary data structure (inverted index) that maps descriptors to objects across multiple images. This intermediary enables efficient correlation by serving as a bridge between objects captured by different recording devices with disparate fields of view, allowing the system to match objects based on their descriptors rather than direct image comparison.
2Measurement precision
If hierarchical classification and descriptor generation are performed for all detected objects, then object identification accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The system performs preliminary hierarchical classification and descriptor generation for all detected objects across multiple images before attempting correlation. By pre-computing these descriptors and organizing them in an inverted index data structure, the system avoids redundant processing during the correlation phase, significantly reducing overall processing time while maintaining high identification accuracy.
3Adaptability or versatility
If multiple recording devices with different fields of view are used to track objects, then object tracking coverage is improved, but system complexity and data correlation difficulty increase
Solution Approach 1:
The patent creates a universal descriptor framework that works across multiple recording devices with different fields of view, camera types, and perspectives. The hierarchical classification categories and inverted index data structure provide a multi-functional solution that handles diverse input sources uniformly, reducing system complexity by eliminating the need for device-specific correlation algorithms.
Data Source
AI summary
Object detection and classification across disparate fields of view are provided. A first image generated by a first recording device with a first field of view, and a second image generated by a second recording device with a second field of view can be obtained. An object detection component can detect a first object within the first field of view, and a second object within the second field of view. An object classification component can determine first and second level classification categories of the first object. A data processing system can create a data structure indicating a probability identifier for a descriptor of the first object. An object matching component can correlate the first object with the second object based on the descriptor of the first object, the probability identifier for the descriptor of the first object, or a descriptor of the second object.