紙張影像質地分析之研究(I)並生矩陣法

紙張交纖的評估使用最多的方法是應用紙張質量或光學密度的變異係數，即一般所謂交織指數。交織指數評估紙張交織只是以紙張影像的吸光度變異進行之，但對於變異範圍的空間關係無法掌握。本研究探討以二級統計法即並生矩陣法來顯示像素灰階的空間相關性，計算灰階變異的或然率，再導出影像地參數如能量、雜亂度、對比、相關及均勻度等特徵，作爲紙張交織定量評估之依據。本研究結果顯示，藉著並生矩陣導出的特徵參數評估紙張交織，短纖紙漿抄造的紙較長纖者交纖較佳。隨抄紙沉降時間的增加，纖維絮凝趨勢增加，並生矩陣參數顯示交織有惡化情形。試驗比較此二級統計法與吸光度變異法對紙張交織的判別能力，並生矩陣法有較多參數作爲判定，其辨別能力較強並能發揮其對紙張基重變異性空間關係的評估功能。

關鍵字

紙張交織；影像分析；手抄紙；並生矩陣；質地分析

並列摘要

The most frequently used method of paper formation evaluation is by formation index, or coefficient of variation (COV) of mass or optical density variation of the paper. The method, however, tells nothing about the spatial relationship among the variations. This study aims to develop an effective 2nd order statistical method, the cooccurrence matrix approach, capable of providing information on gray tone variation of the pixels with respect to their neighbors. The probabilities of tonal variation were used to calculate parameters such as energy, entropy, correlation, contrast, and homogeneity, etc. These were then used for the evaluation of paper formation. The results suggest that comparable results were obtained with COV of light absorbance and the cooccurrence matrix methods. Comparing the results of light absorbance COV with the cooccurrence matrix suggests that the latter can differentiate more effectively ampmg papers of similar basis weight variations. Textural analysis has the ability of establishing relative spatial variations of the gray levels of the image pixels, and hence is more powerful than the traditional light absorbence method.

並列關鍵字

paper formation ； image analysis ； handsheet ； cooccurrence matrix ； spatial gray level correlation

國際替代計量

紙張影像質地分析之研究(I)並生矩陣法

未授權

主題瀏覽