Video Frame Trimming for Reliable Vehicle Image Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle to satisfactorily extract movable objects, such as vehicles, from frames of a moving image for posting on social networking services.
Innovation Solution
An image processing apparatus that includes a storage unit and a processing unit to select and trim frames based on evaluation values, using a learned model for image recognition, to create an extraction image of a movable object like a vehicle, adjusting trimming ranges to optimize the evaluation value.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional object detection is used on each frame of moving image, then object detection can be performed, but it is difficult to satisfactorily extract movable objects such as vehicles from each frame
Solution Approach 1:
The system performs preliminary actions by detecting the target object in a reference frame before processing individual frames. The detected object position and size information from the reference frame is used to define initial trimming ranges for subsequent frames, enabling more reliable extraction of movable objects throughout the video sequence.
Solution Approach 2:
The system creates a copy of the target object detection result from the reference frame and applies it across multiple frames. By copying the object detection information (position, size) from one frame to define trimming ranges in other frames, the system achieves consistent and reliable object extraction even when the object moves.
2Manufacturing precision
If trimming range is adjusted to improve evaluation value, then image quality improves, but processing complexity increases
Solution Approach 1:
The system changes parameters by adjusting trimming ranges based on evaluation values calculated from multiple factors (object position, size, vividness, brightness, cutoff degree). By systematically varying the trimming range parameters and evaluating their impact, the system optimizes image quality while maintaining manageable processing complexity through automated evaluation.
3Measurement precision
If multiple evaluation items are used to determine trimming range, then extraction accuracy improves, but processing time increases
Solution Approach 1:
The system applies partial action by selectively using multiple evaluation items (object position, size, vividness, brightness, cutoff degree) rather than all possible evaluation criteria. This approach achieves sufficient extraction accuracy by focusing on the most relevant factors while avoiding the time cost of evaluating every possible parameter.
Data Source
AI summary
An image processing apparatus includes: a storage unit that stores a plurality of frames; and a processing unit that selects at least one selection frame from the plurality of frames and that outputs an extraction image extracted from the selected selection frame, wherein the processing unit outputs, as the extraction image, the image of the initial trimming range, and the evaluation value of the image of each trimming range is set based on at least one evaluation item of the following evaluation items: a positional relationship of a target object in the image of the trimming range, a size of the target object in the image of the trimming range, a vividness of the target object in the image of the trimming range, a brightness of the image of the trimming range, and a degree of cutoff of the target object in the image of the trimming range.