Object Detection Method Reducing False Alarms in Security Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video cameras, especially those using AI detection methods, suffer from high false alarm rates when detecting stationary objects due to small position deviations in images, leading to incorrect identification of movement.
Innovation Solution
An object detection method that compares frames taken in different time periods to determine if the object's position variation is within preset thresholds, using a count-based approach to differentiate between stationary and moving objects, and incorporates sensor data to validate the state of the object.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AI detection method is used to detect object movement, then detection sensitivity is improved, but false alarm rate increases for stationary objects
Solution Approach 1:
The patent applies dynamics by transitioning from static single-frame detection to dynamic multi-frame sequential detection. The system captures multiple frames over time and evaluates position variations across the sequence, allowing it to distinguish between temporary position deviations (stationary objects) and sustained position changes (moving objects). This temporal dimension transforms the detection approach from a static snapshot to a dynamic process, resolving the contradiction between sensitivity and false alarms.
Solution Approach 2:
The patent implements feedback mechanisms by continuously comparing object positions across multiple frames and using this information to adjust detection decisions. The system accumulates position variation data over time and uses this feedback to determine whether an object is truly moving or merely exhibiting detection noise. This feedback loop enables the system to maintain high sensitivity while filtering out false alarms through iterative validation.
2Productivity
If single frame detection is used, then detection speed is improved, but detection accuracy deteriorates for stationary objects
Solution Approach 1:
The patent applies preliminary action by capturing and storing multiple frames in advance before making a detection decision. Instead of immediately determining movement from a single frame, the system pre-captures a sequence of frames and prepares position data for subsequent analysis. This preliminary data collection enables more accurate distinction between stationary and moving objects without significantly impacting detection speed, as the comparative analysis can be performed efficiently on the pre-captured data.
Data Source
AI summary
This application discloses an object detection method, a security device, and a readable storage medium. The object detection method includes: obtaining a plurality of first frames acquired by imaging an object within a first period of time; obtaining a plurality of second frames acquired by imaging the object within a second period of time, where an end time of the first period of time is a start time of the second period of time; comparing each of the second frames with the plurality of first frames one by one to determine whether a position variation of the object between the second frame and the first frame is less than a first preset threshold; if it is determined that the position variation of the object is less than the first preset threshold, increasing a count by one (1) and obtaining a total count after comparison with the plurality of second frames is completed; determining whether the total count reaches a second preset threshold; and if the total count reaches the second preset threshold, determining that the object is stationary; or if the total count does not reach the second preset threshold, determining that the object is moving. This application improves accuracy of a detection method such as an AI detection method in detecting a state of a stationary object and reduces false alarms.