论文部分内容阅读
引入生物显微镜用于探测病理切片离焦量。针对反射光斑内部纹理缺陷、畸变严重的特殊情况,通过自适应中值滤波、基于Ostu算法的Canny边缘检测(去除非圆弧边缘部分)、畸变校正、最小二乘法拟合非完整圆等算法,直接计算光斑半径大小及离焦方向,建立光斑半径与离焦量的线性关系模型。实验结果表明,与传统的灰度重心法相比,离焦量计算方法抗噪性强、稳定性高,可大幅度地提高线性关系模型的范围及光斑半径、圆心位置的准确度。
Biological microscopes were introduced to detect the amount of defocused pathology. Aiming at the special cases of the internal texture defects and distortion in the reflected speckle, adaptive median filtering, Canny edge detection based on Ostu algorithm (removing non-arc edge), distortion correction, least square fitting non-perfect circle algorithm, Directly calculate the spot radius size and defocus direction, and establish a linear relation model between spot radius and defocus amount. The experimental results show that compared with the traditional grayscale centroid method, the defocus amount calculation method has strong anti-noise and high stability, which can greatly improve the range of the linear relation model and the accuracy of the spot radius and the center of the circle.