Object Detection Apparatus False Negative Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object detection systems fail to adequately reduce false negatives, where objects intended to be detected are missed, due to the limited performance of the first detector used in the detection process.
Innovation Solution
An object detection apparatus comprising an acquisition unit, a first detector, a false negative determining unit, and a second detector, where the false negative determining unit compares detection results across frames to identify missed detections, and the second detector, differing in performance, is applied to the peripheral area of the first detector's results to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single detector is used for object detection, then the detection process is simple and fast, but false negatives occur frequently due to limited detection performance
Solution Approach 1:
The detection system is segmented into multiple detectors (first detector and second detector) with different performance characteristics. Each detector handles specific detection tasks, and their results are combined to achieve more reliable detection while maintaining system manageability through modular architecture.
Solution Approach 2:
Different detectors are configured with different detection parameters and performance characteristics. The first detector uses one set of parameters optimized for certain conditions, while the second detector uses different parameters optimized for other conditions, allowing the system to adapt to various detection scenarios and reduce false negatives.
2Reliability
If two detectors are used to reduce false positives, then detection reliability improves, but the number of false negatives increases due to the first detector's limited performance
Solution Approach 1:
The system incorporates feedback mechanisms where the determination unit analyzes detection results from multiple detectors and adjusts subsequent detection operations. When false negatives are detected, the system feeds back to re-detect those objects using appropriate detectors, continuously improving detection precision while maintaining false positive reduction.
Solution Approach 2:
The system performs preliminary detection using a first detector that is fast and simple, then prepares for subsequent verification detection using a second detector with different performance characteristics. This preliminary action allows the system to efficiently process most cases while reserving detailed verification for cases that require it, balancing speed and accuracy.
3Productivity
If detection is performed on the whole image using a first detector, then processing speed is maintained, but detection precision for all objects cannot be ensured
Solution Approach 1:
Different regions of the image are processed with different detection qualities. The first detector processes the entire image with standard precision for speed, while the second detector applies enhanced precision to specific regions or objects identified as needing verification. This local quality approach maintains overall processing speed while ensuring high precision where needed.
Solution Approach 2:
Instead of applying maximum detection precision to the entire image (excessive action), the system applies partial action by using a faster, less precise detector for most of the image and reserving enhanced precision for only the necessary portions. This balances processing speed with detection precision by applying the higher precision only where needed.
Data Source
AI summary
According to one embodiment, an object detection apparatus includes an acquisition unit, a first detector, a determination unit, and a second detector. The acquisition unit acquires frames in a time-series manner. The first detector detects a predetermined object in each of the frames. The determination unit stores detection results corresponding to the frames, compares a first detection result corresponding to a first frame of the frames with a second detection result corresponding to a second frame of the frames, and determines whether false negative of the predetermined object exists in the second frame. The second detector detects the predetermined object in the second frames when it is determined that false negative of the predetermined object exists. The second detector differs in performance from the first detector.