一种去除二值文本图像边缘噪声的新方法

提出了一种去除二值文本图像边缘噪声的新方法。该方法包括3个步骤:边缘噪声预处理,边缘噪声检测和边缘噪声去除。将该方法和其它方法作了比较。实验结果表明,该新方法可以更有效地去除边缘噪声,并在实际的文本图像处理系统中取得了较理想的效果。

基于数学形态学的线粒体电镜图像边缘检测算法

分析线粒体电镜图像的特点,对其进行预处理,用线性结构元素进行运算,初步获取图像边缘,接着得到二值图像,再用方形结构元素进行运算,减少图像中的无用信息,通过开运算和闭运算滤波去噪并填充边缘缺口,最后对图像进行骨骼化处理,得到清晰约简的线粒体边缘。实验结果表明,该算法能较好的实现线粒体电镜图像的边缘检测,为后续的目标识别提供了良好的基础。

AN EDGE DETECTOR BA SED ON WAVELET TRANSFORM FOR SAR IMAGE

Edge detecting is a f undamental issue in image analysis and image processing. In this paper, based on wavelet transform, an edge detector of SAR image is studied to serve the image aided navigation. Though wavelet transform is efficient as an edge detector, a l ot of false edges will be extracted from a SAR image, especially in areas of hig h reflectivity, with wavelet transform and a fixed threshold because SAR images are always contaminated by speckle. Therefore, this paper develops an algorithm by using a variable threshold with constant false alarm rates to help suppressin g the speckle in SAR images. Finally, two simulating tests are given and the res ults show that the effect of the multiplicative speckle is controlled and there are few false edges in the gotten edge images. Moreover, owing to the threshold from the wavelet transform coefficients, the algorithm can assure the real time of SAR image aided navigation.