Object Area Detection Using Time-Series Background Accumulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object area detection devices struggle to distinctly detect an overlying object when a new object appears from a stationary object or when another stationary object overlays a stationary object, due to the lack of a specified target object template and issues with continuous differential images.
Innovation Solution
An object area detection system that includes a unit to detect object areas in continuous input images, a unit to identify stationary areas within these images, accumulates this information as time-series background data, and compares it to detect objects within the object area, allowing for the detection of overlying objects without relying on a target object template.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a template image of a target object is used for pattern matching, then the object can be detected, but the system cannot detect new objects that appear from stationary objects or detect overlying stationary objects
Solution Approach 1:
The system performs preliminary actions by detecting and storing background information of stationary objects before the actual object detection. It accumulates background images and their corresponding position information in advance, creating a database that can be used to detect both new objects and overlying objects without requiring template images of the objects themselves.
Solution Approach 2:
Instead of using template images of target objects for pattern matching, the system creates a copy of the background environment (background images) and uses this copy to detect objects. By comparing current images with the accumulated background information, the system can identify both new objects and overlying objects based on their deviation from the stored background patterns.
2Measurement precision
If continuous differential images are used to detect moving objects, then moving objects can be detected, but stationary overlying objects are not detected and single-color moving objects cause false positives
Solution Approach 1:
The system segments the detection process into multiple independent components: background information accumulation, object area detection, stationary area detection, and overlying object detection. Each component handles specific aspects of object detection separately, allowing the system to detect moving objects through differential imaging while simultaneously identifying stationary objects and overlying objects through background comparison, thereby eliminating false positives.
Solution Approach 2:
The system introduces background information as an intermediary element between the current image and the detection process. By comparing current images with accumulated background information, the system can distinguish between actual objects and false patterns, enabling reliable detection of both moving and stationary objects without the false positives caused by single-color moving objects in continuous differential images.
Data Source
AI summary
To enable detection of an overlying object distinctively even if a stationary object is overlaid with another stationary object or a moving object. A data processing device includes a first unit which detects an object area in a plurality of time-series continuous input images, a second unit which detects a stationary area in the object area from the plurality of continuous input images, a third unit which stores information of the stationary area as time-series background information, and a fourth unit which compares the time-series background information with the object area to thereby detect each object included in the object area.