taiwan-license-plate-recognition-research-tlprr Computer Vision Project
Updated 10 months ago
7.4k
390
Metrics
English version
Taiwan license plate recognition research
- The license plate recognition project is specially designed for Taiwan license plates. It can identify most of the white plates of automobiles and motorcycles. As for license plates of other colors, there are fewer data sets, so it may not be able to accurately identify them.
- The data is mainly from the data of Roboflow universe, but in a certain data cleaning, I accidentally lost the relevant links
Related projects(and more...)
https://universe.roboflow.com/master-j9s6d/dl_final https://universe.roboflow.com/st-bug8p/plate-pdm1y https://universe.roboflow.com/class-3icb6/license-plate-wu0bx https://universe.roboflow.com/project-oee82/license-bha52
Dataset highlight
- There are sufficient image data for white-brand automobiles and motorcycles
- Image cropped into square(using Python3 and OpenCV to crop images)
- It can be used to detect license plates of driving recorders and moving objects, close range and when the license plate can be captured by the driving recorder lens
If you have other questions or want to discuss this data set, you can contact: https://t.me/jtx257
中文版本
台灣車牌識別研究
- 車牌識別項目專為台灣車牌設計。可以識別大部分汽車、摩托車的白板。至於其他顏色的車牌,數據集較少,可能無法準確識別。
- 數據主要來自Roboflow universe的數據,但在某次數據清洗中,不小心丟失了相關專案的連結
相關項目(還有更多等待找回)
https://universe.roboflow.com/master-j9s6d/dl_final https://universe.roboflow.com/st-bug8p/plate-pdm1y https://universe.roboflow.com/class-3icb6/license-plate-wu0bx https://universe.roboflow.com/project-oee82/license-bha52
數據集特點
- 白牌汽車和摩托車有足夠的圖像數據
- 圖片裁剪成正方形(使用python3和opencv裁剪圖片)
- 可用於偵測行車紀錄器及移動目標的車牌,近距離以及在行車紀錄器鏡頭可拍攝到車牌的情況下
如果對此資料集有其他疑問或想討論的,可聯繫: https://t.me/jtx257
備註
-
如果你正在嘗試著需要更高準確度的商業車牌字元辨識,或許你可以考慮該教授(https://blog.udn.com/yccsonar/article) 匯總了有關車牌辨識系統的相關資料,特別是一位教授提出的非AI/ML方法。
-
參考文獻
-
教授主張其開發的車牌辨識系統不依賴人工智慧(AI)或機器學習(ML),並認為使用ML來進行物件辨識是一種資源的浪費和效率低下。該系統的準確度據稱超過99%。參考資料: 網頁存檔。
-
該教授自稱其開發的車牌辨識系統是台灣市場的主流產品。參考資料: 網頁存檔。
-
在另一篇文章中,該教授表達了對使用人工智慧和機器學習技術的輕蔑。參考資料: 網頁存檔。
- 的確有些系統是連AI、Machine Learning、Deep Learning都用不著的,但這也擋不住人們去嘗試用在別的地方。
- 普通的人們顯然具有某種程度的"懶惰",如果有一個輪子,功能全面,而教育告訴他們: "不要重複製造輪子",我想會自己去嘗試那些從磚頭開始,自行從"磚塊",建設成"房子"的人可為是少數,但不可能否能,這樣確實在某些時候是有用的,像是執行效率上。
- 試想一個情境,通常你會去書局買剪刀去剪紙,還是你會去拿鐵塊融化鑄造出一把剪刀?哪一些公司會願意讓員工如此嘗試?
- OK,以上其實都是loser的發言,教授確實厲害,respect!
- 依然是尊重教授的選擇,只是希望教授的文章少點火氣,但要怎麼發我確實管不著。
Use This Trained Model
Try it in your browser, or deploy via our Hosted Inference API and other deployment methods.
Build Computer Vision Applications Faster with Supervision
Visualize and process your model results with our reusable computer vision tools.
Cite This Project
If you use this dataset in a research paper, please cite it using the following BibTeX:
@misc{
taiwan-license-plate-recognition-research-tlprr_dataset,
title = { taiwan-license-plate-recognition-research-tlprr Dataset },
type = { Open Source Dataset },
author = { JackResearch0 },
howpublished = { \url{ https://universe.roboflow.com/jackresearch0/taiwan-license-plate-recognition-research-tlprr } },
url = { https://universe.roboflow.com/jackresearch0/taiwan-license-plate-recognition-research-tlprr },
journal = { Roboflow Universe },
publisher = { Roboflow },
year = { 2024 },
month = { jan },
note = { visited on 2024-11-22 },
}