Parallel Pipeline Architecture for License Plate Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing license plate identification systems face inefficiencies due to sequential processing of image processing and identification tasks, which can lead to extended processing times and fail to meet user requirements.
Innovation Solution
The proposed method and system implement a parallelized architecture by configuring processors to sequentially obtain images, decompose them into vehicle and license plate images, and process these through distinct pipeline architectures simultaneously, allowing for concurrent processing of multiple images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sequential processing is used for image processing and identification tasks, then processing steps can be completed in order, but processing time is extended and efficiency is reduced
Solution Approach 1:
The patent segments the license plate identification process into distinct pipeline stages: image acquisition, image decomposition, vehicle detection, vehicle metadata identification, license plate identification, and result merging. Each stage is independently processed through separate processing units, enabling parallel execution of multiple images simultaneously while maintaining the required sequential order within each pipeline stage.
Solution Approach 2:
The patent transitions from one-dimensional sequential processing to multi-dimensional parallel processing by implementing multiple pipeline architectures that can execute simultaneously. Different images are processed in parallel across multiple pipeline stages, effectively adding a temporal dimension to the processing architecture while preserving the sequential integrity of individual processing streams.
2Productivity
If parallel processing architecture is implemented, then processing efficiency is improved and computing resources are maximized, but not all tasks can be executed in parallel due to sequential requirements
Solution Approach 1:
The patent divides the processing system into multiple independent pipeline stages, each handling specific tasks. This segmentation allows the system to implement parallel processing where possible while maintaining sequential processing where required, thereby managing architectural complexity through modular design while still achieving improved processing efficiency.
3Measurement precision
If multiple processing stages are implemented for vehicle detection and license plate identification, then identification accuracy is improved, but processing time is extended
Solution Approach 1:
The patent ensures continuous processing by implementing pipeline architectures where multiple images are processed simultaneously through different stages. While each individual image undergoes multiple processing stages for accurate identification, the pipeline structure allows the system to maintain continuous useful action by processing different images in parallel, thereby preventing idle time between stages and maintaining high identification accuracy without excessive delay.
Data Source
AI summary
An identification method and an identification system for a license plate are provided. The method includes: obtaining images in sequence; decomposing each image into a vehicle image and a license plate image, inputting the vehicle image into a vehicle detection model to detect a vehicle through first processing stages; inputting the detected vehicle into a vehicle metadata identification model to obtain vehicle metadata through second processing stages; inputting the license plate image into the license plate recognition model to identify license plate information through third processing stages; and merging and processing the license plate information and vehicle metadata to generate a license plate recognition result. The first processing stages and the second processing stages form a first pipeline architecture, the third processing stages form a second pipeline architecture, and the first pipeline architecture and the second pipeline architecture are executed simultaneously.