模糊演化式影像雜訊濾波器於演化式硬體環境下的設計 = The Desig...
國立高雄大學電機工程學系碩士班

 

  • 模糊演化式影像雜訊濾波器於演化式硬體環境下的設計 = The Design of Fuzzy Image Filters for Evolvable Hardware Environments
  • 紀錄類型: 書目-語言資料,印刷品 : 單行本
    並列題名: The Design of Fuzzy Image Filters for Evolvable Hardware Environments
    作者: 陳建榮,
    其他團體作者: 國立高雄大學
    出版地: [高雄市]
    出版者: 撰者;
    出版年: 2012[民101]
    面頁冊數: 125面圖,表格 : 30公分;
    標題: 去除雜訊
    標題: denoising
    電子資源: http://handle.ncl.edu.tw/11296/ndltd/48065558285496440480
    附註: 104年10月31日公開
    附註: 含參考書目
    摘要註: 雜訊過濾(noise filtering)或去噪(denoising)指的是將訊號中的雜訊減少或是移除,是影像或聲音處理過程中的重要方法。一般去除雜訊的方法如果在預先知道雜訊的類型,通常有不錯的過濾效果;但是在未知雜訊類型的情形下,則缺乏了適應性(adaptability)與靈活性(flexibility),往往過濾效果不佳。演化式硬體(evolvable hardware)結合演化式演算法(evolutionary algorithms) 及可重構式(reconfigurable)硬體可以依照外在環境的變化,以演化的方式調整內部結構,獲得具有適應性與靈活性的輸出。本論文以演化式硬體為平台,設計具彈性的影像雜訊過濾器。本研究根據像素之間的相似度(similarity)與差異性(divergence)建立雜訊型態模型,並藉由模糊推論進行雜訊型態的判定,建造出對應的演化式硬體影像雜訊濾波器。本論文設計演化式硬體的演化機制與模糊分析模型,並與其他過濾雜訊的方法進行實驗的比較與分析。實驗結果顯示此方法所訓練出來的濾波器具備更好過濾雜訊的能力。 Denoising is to remove or reduce noises from the contaminated images and is an important research area in image processing. Denoising methods can work well if the noise models can be known in prior. However, it may lack of flexibility andadaptability when un-modeled types of noises are encountered. Evolvable Hardware(EHW) is a combination of evolutionary algorithm and reconfigurable hardware devices. EHW can change its architecture adaptively and produce flexible results for un-modeled problems. This study analyzes types of noises by fuzzy rules and builds EHW models for filtering image with various types of noises. Noise types are categorized and defined in fuzzy terms according to the similarity and divergence of the masked pixels in a sliding window. The value for restoring the noisy pixel is calculated from the outputs from the EHW models. With the proposed method, the efficiency of training EHW models and accuracy of image filtering are both improved. This thesis evaluates and compares the performance of the proposed method with other ones.
評論
Export
取書館別
 
 
變更密碼
登入