Event-Based Sensor Object Detection for Low-Power Applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection methods using traditional imaging devices are energy-intensive and inefficient, making them unsuitable for quick and accurate detection of moving objects, particularly in portable devices.
Innovation Solution
An object detection method utilizing an event-based sensor to generate an event image, determine feature vectors from target and neighbor pixels, and classify objects using a trained classifier, which enables efficient and rapid identification of moving objects with reduced power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a typical imaging device based on CCD or CMOS is used to detect motion by obtaining images and classifying motion sections, then the moving object can be identified, but the detection requires a large amount of time and cannot quickly trace the moving object
Solution Approach 1:
The patent changes the fundamental parameter of image acquisition from frame-based capture to event-based sampling. Instead of capturing complete images at fixed intervals and then processing them to detect motion, the system directly samples pixel changes asynchronously as they occur. This parameter change transforms the detection process from a time-intensive frame analysis approach to a real-time event response approach, achieving both high accuracy and fast detection speed.
2Speed
If the imaging device is kept open to quickly capture a moving object, then the object can be captured rapidly, but a relatively large amount of energy is consumed
Solution Approach 1:
The patent implements periodic sampling only when necessary - specifically, when pixel value changes are detected. The event-based sensor operates in a periodic manner by continuously monitoring pixel states and triggering image capture only during events when motion is detected, rather than maintaining continuous operation. This allows rapid object capture while dramatically reducing energy consumption compared to keeping the imaging device continuously active.
3Loss of information
If traditional frame-based image processing is used to identify moving objects, then complete image information is obtained, but the processing requires significant computational resources and time
Solution Approach 1:
The patent extracts only the essential information needed for motion detection - specifically, the changed pixel values and their coordinates - rather than processing complete frame images. By taking out only the relevant event data (pixel changes) from the full image stream, the system maintains sufficient information for accurate object identification while dramatically reducing computational burden and improving processing efficiency.
Data Source
AI summary
A method and apparatus for detecting an object using an event-based sensor is provided. An object detection method includes determining a feature vector based on target pixels and neighbor pixels included in an event image, and determining a target object corresponding to the target pixels based on the feature vector.