基于最大类间距的生物芯片图像网格定位算法

Gridding of Microarray Image by the Maximum Between-Class Varian

为了解决在生物芯片图像网格化过程中网格线冗余、缺失及出现在信号点上的现象,提出一种结合最大内间距(Otsu)和改进了网格校准方法的网格定位算法。首先将调整后的最大内间距作为最佳阈值得到生物芯片图像的二值参考图像,确定初始网格线;其次对初始网格线进行测试、校准;最后获得优化后的网格线,得到网格化图像。实验结果表明,改进后算法的网格线准确率高于其它算法。改进后算法简单实用,能实现生物芯片图像网格化功能,可大大减小后续图像分析、处理的复杂程度。

Aiming at the phenomenons that the grid lines are redundant, missing, and appearing on the signal points in the process of gridding microarray images, we put forward a gridding algorithm which was based on the Maximum Between-Class Variance （Otsu） and improved grid lines optimizing method. First the adjusted maximize between-class variance was used as the optimal threshold to gain the binary reference signal of microarray image and determin the initial grid lines ; n. Next the initial grid lines were tested and calibrated. Last the optimized grid lines and gridding of microarray image were obtained. The experimental results show that the accuracy of the improved algorithm is significantly higher than other algorithms. In summary, the improved algorithm is simple and can accomplish the function of image gridding, which greatly reduced the complexity level of subsequent image analysis and processing.