Moving Body Detection Using Motion-Compensated Approximate Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving body detection technologies face challenges in accurately tracking and precisely detecting moving bodies using cameras with short or long exposure times, as they either produce discontinuous images leading to low precision or widen the search range, increasing occlusion and reducing detection accuracy.
Innovation Solution
A moving body detecting device that acquires motion information from a camera capable of alternating between two exposure times, generates approximate images by convolving blur filters derived from camera motion into one image set, and detects moving bodies by finding differences between these images, effectively isolating and removing camera motion blur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a camera with short exposure time (high frame rate) is used, then the searching range for feature points is narrowed down, but the information on moving body becomes discontinuous and detection precision is insufficient
Solution Approach 1:
The system performs preliminary action by generating an approximate image from the long exposure time image before comparing it with the short exposure time image. This preliminary approximate image serves as a reference that incorporates motion information, allowing the system to detect moving bodies more accurately while maintaining information continuity across different exposure times
Solution Approach 2:
The approximate image acts as an intermediary between the long exposure time image and the short exposure time image. By comparing the short exposure time image with this intermediate approximate image (rather than directly with the long exposure time image), the system bridges the information gap caused by different exposure times while maintaining detection precision
2Area of stationary object
If a camera with long exposure time (low frame rate) is used, then the moving distance of subject or camera increases and searching range is widened, but detection precision gets lower and occlusion occurs
Solution Approach 1:
The system generates an approximate image as a preliminary reference from the long exposure time image before performing moving body detection. This approximate image contains motion information that helps maintain detection precision even when the searching range is widened due to longer exposure time
Solution Approach 2:
The system dynamically adjusts the comparison process by using motion information derived from the long exposure time image to create an approximate image. This dynamic approach allows the system to adapt to the widened searching range while maintaining detection precision through motion-compensated comparison
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances precision in detecting moving bodies by canceling camera motion blur and isolating subject motion blur, allowing for accurate tracking and improved detection accuracy.
Implementation Method 1
generates approximate images by convolving blur filters derived from camera motion into one image set
Data Source
AI summary
A moving body detecting device in accordance with embodiments includes a motion information acquirer, approximate image generator, and moving body detector. The motion information acquirer generates motion information on an imaging device capable of performing imaging with a first exposure time and a second exposure time longer than the first exposure time. The approximate image generator generates an approximate image generated by approximating one of a first photographic image imaged with the first exposure time and a second photographic image imaged with the second exposure time to another photographic image, on the basis of the motion information. The moving body detecting device detects a moving body captured in the first and second photographic images, on the basis of the approximate image and the other photographic image.