Colored Moving Element Detection Through Motion-Color ROI Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image classification methods for embedded systems face challenges in optimizing processing power without compromising classification performance, particularly in low-cost devices, leading to high computational demands and reduced accuracy due to unnecessary CNN model inferences and resource constraints.
Innovation Solution
A method involving pre-processing video frames to detect motion and color patterns, generating regions of interest, and using digital zoom to focus on the intersection area for CNN inference, reducing the image scope to match model input requirements, thereby optimizing processing and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If CNN model inference is run continuously on all video frames, then detection coverage is improved, but processing power consumption increases and false positives occur during waiting periods
Solution Approach 1:
The patent applies preliminary action by performing pre-processing operations (motion detection, color signature detection, region of interest identification) before CNN inference. This prepares the data in advance and filters out non-relevant frames, so that CNN only processes frames with actual detection potential, reducing unnecessary computations and power consumption while maintaining detection coverage
Solution Approach 2:
The patent extracts only the relevant portions of video frames for CNN processing. By identifying regions of interest through motion and color detection, and applying digital zoom to focus on specific areas, the system extracts only the necessary image data for inference, reducing the computational load on the CNN model while maintaining detection accuracy
2Measurement precision
If full video frames are processed by CNN model, then detection accuracy is maintained, but processing time and computational resources increase
Solution Approach 1:
The patent extracts only the relevant regions from full video frames for CNN processing. By identifying regions of interest through motion and color detection, and applying digital zoom to focus on specific areas, the system extracts only the necessary image data for inference, reducing the computational load and processing time while maintaining detection accuracy on the target elements
Solution Approach 2:
The patent segments the video processing pipeline into distinct stages: pre-processing (motion detection, color detection, ROI identification), digital zooming, and CNN inference. This segmentation allows each stage to handle specific tasks efficiently, with pre-processing filtering out irrelevant data before it reaches the CNN model, reducing overall processing time while maintaining accuracy
3Productivity
If multiple processor cores are used to run CNN concurrently, then processing speed improves, but hardware complexity and cost increase
Solution Approach 1:
The patent applies preliminary action by performing computationally intensive pre-processing operations (motion detection, color signature detection, region identification) before CNN inference. This reduces the amount of data that needs to be processed by the CNN model, allowing single-core or low-core processors to achieve acceptable processing speeds without requiring multiple expensive processor cores
Solution Approach 2:
The patent extracts only the essential features and regions of interest before CNN processing. By reducing the input data size through digital zoom and ROI selection, the system reduces the computational burden on the processor, enabling efficient operation on embedded systems with limited processing cores while maintaining productivity
Data Source
AI summary
A method of detecting moving elements in a video file, comprising: receiving a video file in which a target colored moving element can be located; extracting video frames into images; pre-process the images. The pre-processing the images comprises detecting a desired motion pattern and generating a desired motion pattern region of interest where said desired motion pattern is detected; detecting a desired color signature and generating a desired color signature region of interest where said desired color signature is detected; and detecting an intersection between the desired motion pattern region of interest and the desired color signature region of interest. The pre-processing the includes images digital zooming on intersection; and inferring, at the intersection, by a convolutional neural network, the target colored moving element in the images and reporting the identification of said target colored moving element.