細(xì)胞的形狀極其不規(guī)則，大小差異明顯，而且采集的切片顯微圖像有細(xì)胞重疊現(xiàn)象，
有一定的噪聲，造成了細(xì)胞面積描述的困難，因此研究全體細(xì)胞的平均面積比研究單個細(xì)胞的面積更有意義。對于細(xì)胞面積的近似計算提出了基于閾值分割和區(qū)域生長算法統(tǒng)計分析細(xì)胞的平均面積的方法。實驗表明此方法能有效表示細(xì)胞面積的平均中心趨勢。
關(guān)鍵詞：圖像分割，區(qū)域生長，細(xì)胞面積，Java
Abstract: It is difficult to describe the area of cell because the shape and size of the cell is
different from each other. More over, The Microscope image of slice has cells lapped over
phenomena and noise. Therefore, it is more significant to analyze the areas of all cells than the
area of single one. This paper proposes a hybrid method based on threshold segmentation and region growing algorithm to estimate the approximate area of cell. The experiment shows that this method is an efficient measurement of the central tendency of the area of cell.
Key Words: Image Segmentation, Region Growing, Area of Cell, Java