Infrastructure Camera Target Recognition with Selective Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing target recognition systems using infrastructure cameras face high processing loads due to complex tracking processes, particularly when multiple targets are detected, leading to false detections and increased computational resources, which can cause unnecessary operations and user anxiety.
Innovation Solution
A target recognition system that classifies detected targets as movable or stationary, applying a tracking process only to movable targets and a simplified determination process to stationary targets, thereby reducing processing load and false detections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a tracking process is applied to all detected targets to reduce false detections, then the reliability of target recognition is improved, but the processing load and computational resource consumption increase
Solution Approach 1:
The patent applies different processing methods to different target classes. Movable targets undergo full tracking processes to ensure reliability, while stationary targets use simplified detection without tracking, reducing overall processing load. This local differentiation optimizes the balance between reliability and productivity.
Solution Approach 2:
The patent segments detected targets into movable and stationary categories based on their characteristics. By dividing the processing workload into two distinct streams with different processing intensities, the system reduces computational resource consumption while maintaining high reliability for movable target recognition.
2Reliability
If a tracking process is applied to stationary targets, then false detections are reduced, but unnecessary operations and computational resources are consumed
Solution Approach 1:
The patent extracts the tracking process from the general processing pipeline and applies it selectively only to movable targets. Stationary targets are processed through a simplified detection path, removing unnecessary computational operations and reducing energy consumption while maintaining adequate reliability.
Solution Approach 2:
The patent applies partial tracking action only where necessary (for movable targets) rather than excessive tracking for all targets. This selective application avoids wasting computational resources on stationary targets that do not require tracking to confirm their presence.
Data Source
AI summary
A target recognition system recognizes a target shown in an image captured by an infrastructure camera. The target recognition system detects a first target shown in the image as a temporary target. When a class of the temporary target is a movable target, the target recognition system applies a tracking process to the temporary target and determines that the first target is a real movable target based on a result of detection of the temporary target in a first period. When the class of the temporary target is a stationary target, the target recognition system determines that the first target is a real stationary target based on a result of detection of the temporary target in a second period without applying the tracking process to the temporary target.