论文部分内容阅读
锐化是图像增强巾一项关键性的技术,但如果图像中包含噪声,噪声也会囚为锐化而放大,最终导致信噪比的降低。探索了一种算法既可以对图像进行锐化滤波,又不降低图像的信噪比。采用模式识别的相关理论,基于隶属度和概率松弛技术对红外图像中由真实边缘和由各种噪声引起的亮度数值变化进行区分,对不同区域采用不同的锐化处理。该算法不同于传统图像锐化算法只基于局部对比度的缺点,在图像锐化过程中考虑图像边缘和噪声的空间分布的差异,改善了传统边缘增强算法对噪声放大的缺点。实验数据表明,该锐化方法未引起信噪比的降低,具有良好的前景和实用价值。
Sharpening is a key technique for image enhancement, but if the image contains noise, the noise is also sharpened and enlarged, eventually leading to a reduction in signal-to-noise ratio. An algorithm is explored to both sharpen and filter the image without reducing the signal-to-noise ratio of the image. Based on the theory of pattern recognition, based on the membership degree and probability relaxation technique, the change of the brightness value caused by the real edge and various noises in the infrared image is distinguished, and the sharpening is applied to different regions. This algorithm is different from the traditional image sharpening algorithm only based on the local contrast shortcomings. In the process of image sharpening, the difference of spatial distribution of image edge and noise is considered, which improves the shortcoming of traditional edge enhancement algorithm for noise amplification. Experimental data show that the sharpening method does not cause a reduction in signal-to-noise ratio, and has good prospects and practical value.