透過您的圖書館登入
IP:18.224.44.108
  • 學位論文

具影像增強之嵌入式系統設計

Embedded System Platform Design With Image Enhancement

指導教授 : 謝景棠

摘要


近年來,影像處理技術( Image Processing Technology) 逐步的實現及應用在人們生活中,然而之前部份影像的研究是先行透過 X86 系統運算,再將指令傳送至目標物動作,如此的設計不但高成本也耗費空間。嵌入式產品成本及體積都低於個人電腦,也因為擁有這些優點讓嵌入式產品近來無論在硬體或軟體上有了相當程度的進展,如智慧型手機 (Smart Phone ) ; 電子書( E-Book )甚至當今熱銷的 I-Phone ; I-Pad 都已不在是 X86 的架構,取而代之的是以SOC為基礎的嵌入式系統設計。 強健的影像處理結果能讓辨識率準確度提高,進而提昇商品的附加價值,例如相機可以清楚辨識人臉表情,自動按下快門抓取最佳時刻,自動販賣機可以辨識年齡,手機可以辨識指紋 ..等等,基於成本及體積的考量將影像強化(Image Enhancement)處理實現在嵌入式系統裡取代原先的X86,會是這幾年相當重要的進展,嵌入式系統成本低,佈線及外觀空間小符合獨立運作,適合各樣消費型產品的開發。 本研究提出一具研究與應用價值的影像強化系統,使用三星 ARM 系列的s3c6410 系統實現影像強化,自動篩選出高動態範圍影像作補償,以得到較佳的影像,實驗證明可以應用在影像系統中。最後引用結構相似性指標( SSIM ),兼顧了影像之亮度、對比度及結構性,綜合性的測試出影像之整體品質。

並列摘要


In recent years,image processing technology is gradually used and realized in our life. However, lots of researches are performed by X86 system to pass result or command to control devices after a basic operation.This design expends not only high cost but also large layout area. The cost of embedded system is low and the space is small.For example,the most popular I-Phone , I-PAD use embedded system design. This is because strong image quality is an important factor with effective recognition result and the product with high level function will also improve extra value,such as facial expression recognition on camera and age recognition on vending machine. Based on the cost and dimension and stand-alone operation , the consideration to use embedded system instead of x86 will be the popular topic in this moment. In this paper , an image enhancement algorithm is proposed performed well on HDR images , and can be realized on the embedded system .

並列關鍵字

Embedded System Image enhancement HDR

參考文獻


[1] 李政憲,公共空間人潮的計數研究,大同大學資訊工程研究所碩士論文,民國九十七年六月。
[4] C. Wong , D. Kortenkamp and M. Speich , A Mobile Robot That Recognizes People ,IEEE International Conference on Tools with Artificial Intelligence, 1995.
[5] 李承翰,雷射測距於文公尺整合性之研究,國立虎尾科技大學光電與材料科技研究所碩士論文,民國九十九年六月。
[9] 陳金徽,數位影像輻射品質之客觀指標評估,國立成功大學
[11] J. Duan and G. Qiu, “Fast Tone Mapping for High Dynamic Range Images,” ICPR 2004. Proceedings of the 17th International Conference on Pattern Recognition, Aug. 2004.

延伸閱讀