Publications-NSC Projects

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 關於區域影像特徵描述子的建構框架
作者 廖文宏
貢獻者 資科系
關鍵詞 影像特徵描述子; 點群排序關係; 影像比對
feature descriptor; local intensity order relations; image recognition
日期 2016
上傳時間 17-May-2017 15:26:04 (UTC+8)
摘要 隨著科技的進步以及網際網路的普及,影像資訊的傳遞已經漸漸取代文字的表達,人們對於影像的需求也越來越多元,使得影像處理技術以及影像資訊分析也就越來越重要。然而,影像中其中一項重要的資訊為特徵描述子,強而有力的描述子能使得影像在辨識、分類等應用上有較佳的回饋,描述子的建構方式根據編碼原則分為:基於區域梯度統計、基於點對關係以及基於點群關係。其中,基於點群關係的編碼方式因為點群的選取及排序過程中,可能會產生過多的關係表示方法數,以至於不利於計算,因此過去較少有利用點群關係的編碼方式所建構而成的特徵描述子。本研究提出描述子建構方式-LIOR,是以點群排序關係為基礎的編碼方式,相較於LIOP方法隨著點群內的點數增加,元素關係數大幅度的成長,造成描述子維度過大,計算時間和空間皆可能需要大量的消耗,而本研究方法足以改善計算維度的問題,重新定義點群關係的排名機制,並以像素值為基準加入權重分配,以區別加權排序之間不同大小差值所造成的影響程度。實驗結果顯示本研究方法對於不同影像劣化效果的資料集,不僅能提升選取多點為一群的影像比對評估效能,同時也能改善點群內元素關係過多的排名表示法,降低以多點為群集的特徵描述子維度,節省了影像比對的計算時間以及空間,仍可維持整體影像配對之效能。
With the advances of imaging technology and the popularity of mobile Internet, images have played an increasingly important role in interpersonal communication. As such, algorithms for automatic classification and recognition of images are being actively pursued by many researchers in the area of computer vision. Robust image features are essential in building effective image recognition engines. These features can be constructed according to various principles, such the distribution of local gradients (Histogram of Oriented Gradients, HOG), the relationship
between two pixels (Local Binary Descriptors, LBD), or local intensity order statistics (Local Intensity Order Patterns, LIOP). Because the feature dimension grows quickly as we consider the ordering relations of a group of N (N>2) pixels, few researchers have exploited local order
statistics among a pixel set to define an image feature. In this research, we propose a novel approach to construct a feature descriptor using local intensity order relations (LIOR) in a pixel group. In contrast to LIOP where the feature dimension increases drastically with the number of
elements in a set, the size of LIOR is manageable. Moreover, LIOR ensures the stability of ordering by
encoding the intensity differences as weights. Two different strategies for assigning the weights have been devised and tested. Experimental results indicate that the proposed methods yield better or comparable performance for different types of image degradation when compared to the original LIOP. Additionally, the storage requirement is significantly lower when the number of pixels in a group increases.
關聯 MOST 104-2221-E-004-009
資料類型 report
dc.contributor 資科系
dc.creator (作者) 廖文宏zh_TW
dc.date (日期) 2016
dc.date.accessioned 17-May-2017 15:26:04 (UTC+8)-
dc.date.available 17-May-2017 15:26:04 (UTC+8)-
dc.date.issued (上傳時間) 17-May-2017 15:26:04 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/109686-
dc.description.abstract (摘要) 隨著科技的進步以及網際網路的普及,影像資訊的傳遞已經漸漸取代文字的表達,人們對於影像的需求也越來越多元,使得影像處理技術以及影像資訊分析也就越來越重要。然而,影像中其中一項重要的資訊為特徵描述子,強而有力的描述子能使得影像在辨識、分類等應用上有較佳的回饋,描述子的建構方式根據編碼原則分為:基於區域梯度統計、基於點對關係以及基於點群關係。其中,基於點群關係的編碼方式因為點群的選取及排序過程中,可能會產生過多的關係表示方法數,以至於不利於計算,因此過去較少有利用點群關係的編碼方式所建構而成的特徵描述子。本研究提出描述子建構方式-LIOR,是以點群排序關係為基礎的編碼方式,相較於LIOP方法隨著點群內的點數增加,元素關係數大幅度的成長,造成描述子維度過大,計算時間和空間皆可能需要大量的消耗,而本研究方法足以改善計算維度的問題,重新定義點群關係的排名機制,並以像素值為基準加入權重分配,以區別加權排序之間不同大小差值所造成的影響程度。實驗結果顯示本研究方法對於不同影像劣化效果的資料集,不僅能提升選取多點為一群的影像比對評估效能,同時也能改善點群內元素關係過多的排名表示法,降低以多點為群集的特徵描述子維度,節省了影像比對的計算時間以及空間,仍可維持整體影像配對之效能。
dc.description.abstract (摘要) With the advances of imaging technology and the popularity of mobile Internet, images have played an increasingly important role in interpersonal communication. As such, algorithms for automatic classification and recognition of images are being actively pursued by many researchers in the area of computer vision. Robust image features are essential in building effective image recognition engines. These features can be constructed according to various principles, such the distribution of local gradients (Histogram of Oriented Gradients, HOG), the relationship
between two pixels (Local Binary Descriptors, LBD), or local intensity order statistics (Local Intensity Order Patterns, LIOP). Because the feature dimension grows quickly as we consider the ordering relations of a group of N (N>2) pixels, few researchers have exploited local order
statistics among a pixel set to define an image feature. In this research, we propose a novel approach to construct a feature descriptor using local intensity order relations (LIOR) in a pixel group. In contrast to LIOP where the feature dimension increases drastically with the number of
elements in a set, the size of LIOR is manageable. Moreover, LIOR ensures the stability of ordering by
encoding the intensity differences as weights. Two different strategies for assigning the weights have been devised and tested. Experimental results indicate that the proposed methods yield better or comparable performance for different types of image degradation when compared to the original LIOP. Additionally, the storage requirement is significantly lower when the number of pixels in a group increases.
dc.format.extent 3139104 bytes-
dc.format.mimetype application/pdf-
dc.relation (關聯) MOST 104-2221-E-004-009
dc.subject (關鍵詞) 影像特徵描述子; 點群排序關係; 影像比對
dc.subject (關鍵詞) feature descriptor; local intensity order relations; image recognition
dc.title (題名) 關於區域影像特徵描述子的建構框架zh_TW
dc.type (資料類型) report