本文提出了一种新颖的图像目标面积测量方法,它适用于具有光滑且不规则形状边缘的目标面积的精确计算。该方法应用动态轮廓线模型,分为2个步骤。首先利用特征搜索和迭代算法使动态轮廓线由初始位置向目标轮廓边缘收敛。其次,对于收敛后的动态轮廓线,使用基于B样条曲线的封闭曲线面积计算公式,准确计算出目标面积。实验表明,该方法的平均计算误差为±0.5%。与传统面积测量方法相比,这种方法具有很强的鲁棒性,它特别适用于目标背景存在各种噪声和目标内部具有复杂边缘的面积计算情况。
A novel approach to the area measurement of an image target is proposed. This approach is suitable for exact calculating the area of a target with smooth and irregular edges. A dynamic contour model was applied and this approach consists of two steps. Firstly, the dynamic contour converges to the target contour edge from its initial position through feature searching and iteration algorithm. Secondly, an area calculation formula based on B-spline closed curve is applied to compute the target area exactly. Experiments indicate that the average calculation error for this approach decreases to ±0.5%. Compared with conventional area measurement approaches, this approach is robust and suitable for the area computation of the image target with background noise or many complicated internal edges.