基于红外图像的GVF Snake轮廓提取算法的研究

针对红外图像目标具有边界模糊不清，区分效果较差的缺点，结合Ostu阈值法和梯度矢量流主动轮廓模型（GVF Snake），提出一种目标轮廓自动提取方法。采用Ostu法先对图像进行分割，然后将得到的边界作为Snake模型的初始边缘轮廓,利用 GVF Snake特性将初始轮廓准确地收敛到目标边界。由于Ostu算法具有将目标物体从复杂背景中分割开来的优点，使得在应用 GVF Snake模型对复杂图像进行分割时减少了人工的干预。实验证明：该方法运算速度快，能够快速地收敛到目标轮廓，并准确地跟踪目标，具有一定的抗噪能力。

Algorithm of GVF Snake contour extraction for infrared image

To overcome the blurred borders and poor distinction of the infrared images, an automatic object contour extraction method based on the Ostu threshold method and GVF Snake is proposed. The Ostu algorithm is adopted to segment the image. Then the extracted object contour is taken as an initial contour of Snake model for precise segmentation computation. After that, the initial contour is converged to the target borders precisely by utilizing the feature of GVF Snake. With GVF Snake model, the manual operation is reduced during the segmentation of complex images thanks to the advantages of the Ostu algorithm. Experiments show that such algorithm is fast in calculation, it can converge the contour to the target borders rapidly and track a target precisely, and it is immune to noise.