Image Processing Apparatus Adaptive Segmentation Stationary Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies restrict the position of segmented regions, leading to prolonged designation of the same location for objects that remain stationary, failing to adaptively determine the segmented region based on the duration an object is stopped.
Innovation Solution
An image processing system and method that dynamically adjusts the segmented region by using a camera with a region setting unit, object detection unit, and communication unit to detect and switch targets based on evaluation criteria such as stopping time and tracking time, ensuring adaptive segmentation of stationary objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the segmented region is determined based on detection target objects, then the segmentation accuracy is improved, but the position of segmented region becomes restricted and cannot adapt to stationary objects
Solution Approach 1:
The patent introduces dynamic switching between two detection modes: object detection mode for moving objects and stop detection mode for stationary objects. The system automatically switches modes based on whether objects are detected, enabling the segmented region determination to adapt to different object states (moving or stationary) rather than being restricted to a single detection approach
Solution Approach 2:
The patent changes the detection parameters based on the object state. When objects are detected, it uses object detection parameters to determine segmented regions. When no objects are detected (indicating stationary objects), it switches to stop detection parameters, thereby adapting the segmentation criteria to match the actual scene conditions
2Stability of the object's composition
If the same location continues to be set as segmented region for stationary objects, then the detection consistency is maintained, but the processing time is wasted and efficiency decreases
Solution Approach 1:
The patent implements periodic detection with mode switching: it periodically checks for object detection results and switches between object detection mode and stop detection mode. This periodic re-evaluation allows the system to maintain detection consistency for stationary objects while avoiding wasteful continuous processing of the same location, thereby improving overall processing efficiency
3Reliability
If object detection is performed continuously, then the detection reliability is improved, but the computational load increases and energy consumption rises
Solution Approach 1:
The patent dynamically adjusts detection intensity by switching between two modes: object detection mode for active object tracking and stop detection mode for stationary object identification. This dynamic approach maintains detection reliability by appropriately selecting the detection mode based on scene conditions while reducing unnecessary computational load and energy consumption compared to continuous full-scale object detection
Solution Approach 2:
The system changes detection parameters based on the current mode. In object detection mode, it uses parameters optimized for detecting moving objects. In stop detection mode, it switches to parameters optimized for identifying stationary objects, thereby maintaining detection reliability while adapting computational resources to actual needs and reducing overall energy consumption
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus, including circuitry configured to generate or receive a first image of a sequence of images including an object. The circuitry is configured to determine a length of time movement of the object is below a predetermined movement threshold. The circuitry is further configured to identify the object as a target object based on the determined length of time the movement of the object is below the predetermined movement threshold.