生物晶片–基因微陣列是一種新興且能夠同時偵測分析千萬點基因是否有表現的工具。基因微陣列的影像擷取與分析辨識是進行後續實驗分析的第一個步驟，也是不可或缺的重要步驟。分析一張含有萬點基因表現點的高密度基因微陣列影像時，如何將不同形狀或是歪斜位置的基因表現點從基因微陣列影像中正確的切割出來，對於往後資料分析的精確度會直接影響到實驗的最後結果。本研究之重點在於如何藉由理想的影像邊緣偵測演算法(Canny Edge Detection Algorithm)從一張基因微陣列影像中正確的切割出每一個基因表現點的位置。以史丹福大學基因微陣列資料庫中的基因微陣列影像做為本研究的基因表現點切割測試資料，以本研究建立自初步基因表現點方格劃分、偵測基因表現點邊緣、動態修正基因表現點方格功能，測試找尋及分析比較基因微陣列影像上面的基因表現點，證明可提供一個良好的基因微陣列影像分析工具，可提高後續資料分析的精確度。 Biochip – microarray is a new tool used to examine expression levels for thousands of genes simultaneously. In the process of the microarray experiment analysis, image acquisition and analysis is the first step and it is the important process. To analyze a high density microarray image with thousands of gene spots, how to segment the gene spot accurately from irregular shapes and crooked position directly affects the final result of bio-experiment. In this paper, spots the microarray image is focused on how to use an ideal edge detection algorithm – Canny edge detection algorithm - to segment accurately. By testing the microarray images obtained from the Standford Microarray Database. With the system(gridding、spot edge detection、dynamic gridding) built by in paper. The system is verified and proved to be able to provide a good method for each spot segmentation from an in input microarray accurately to raise the data precision.
