Dynamic Image Frame Scaling for Real-Time Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital camera systems face challenges in real-time object tracking due to increased processing time when a target subject is large on the image frame, as the number of pixels forming the region of interest becomes large, leading to longer extraction times.
Innovation Solution
An image processing device that includes an extractor to identify a region of interest, a size decision unit to determine the image size based on the region's size, and an image converter to adjust the image frame size, allowing for efficient extraction of the region of interest by reducing the image frame size when the target subject is large, thereby shortening processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the image frame size is maintained at full resolution, then the extraction precision of the region of interest is improved, but the processing time increases
Solution Approach 1:
The patent divides the image processing into two stages: first, a rough extraction is performed on a down-sampled (reduced resolution) image to quickly identify the region of interest; second, a precise extraction is performed on the original full-resolution image using the coordinates obtained from the rough extraction. This segmentation of processing stages allows the system to benefit from both speed (in the first stage) and precision (in the second stage), resolving the contradiction between extraction precision and processing time.
2Productivity
If the image frame size is reduced, then the processing time is shortened, but the extraction precision deteriorates
Solution Approach 1:
The patent performs a preliminary extraction action on a down-sampled image before the final precise extraction. By first processing a reduced-resolution version of the image to obtain approximate coordinates of the region of interest, the system prepares the groundwork for the subsequent precise extraction on the full-resolution image. This preliminary action enables faster initial processing while setting the stage for high-precision results.
Solution Approach 2:
The down-sampled image serves as an intermediary between the full-resolution image and the final extraction result. Instead of directly extracting from the full-resolution image (which would be time-consuming) or from the down-sampled image alone (which would be imprecise), the system uses the down-sampled image as an intermediate step to obtain coordinate information that guides the precise extraction from the full-resolution image.
Data Source
AI summary
An image processing device includes: an extractor configured to extract a region of interest which satisfies a specified condition from a first image frame; a size decision unit configured to decide an image size according to a size of the region of interest extracted by the extractor; and an image converter configured to change the size of an image frame to be processed according to the image size decided by the size decision unit to generate a transformed image frame. The extractor extracts a region of interest which satisfies the specified condition from the transformed image frame.