CAJ | 학술논문

　　针对最大散度差准则运用到图像分割时存在参数C不确定的问题，借助模糊概念导出参数C的初步计算公式，用二分法的迭代思想自适应地优化参数C，使最大散度差分割法获得的分割阈值相应地优化到最佳。实验证明自适应的最大散度差阈值分割法能使分割后的图像区域更均匀、细节保留得更好。
　　침대최대산도차준칙운용도도상분할시존재삼수C불학정적문제，차조모호개념도출삼수C적초보계산공식，용이분법적질대사상자괄응지우화삼수C，사최대산도차분할법획득적분할역치상응지우화도최가。실험증명자괄응적최대산도차역치분할법능사분할후적도상구역경균균、세절보류득경호。
Aiming at the uncertainty of parameter C in Maximum Scatter Difference(MSD)criterion applied in image segmen-tation, a preliminary formula is derived with fuzzy concept to calculate C, then optimize C adaptively based on dichotomy itera-tive, and the threshold obtained by MSD segmentation accordingly is optimized to a best value. Experimental results show that the Adaptive MSD(AMSD)segmentation can not only make the regions of the segmented image more uniform, but also keep the image details better.