改进的小波变换算法在图像融合中的应用

Application of improved wavelet transform algorithm in image fusion

为了改善传统图像融合方法在精确制导武器系统中对目标检测模糊、识别率低与实时性差等缺陷,采用了一种将小波变换与Canny算子相结合的图像融合的新方法。该方法的具体改进在于首先对源图像在垂直和水平方向上进行了适合图像重构的3层小波分解,并依据各分解层不同频率分量的自身特性,采用独特的融合规则,即对低频分量采用加权平均融合算法,对高频分量采用Canny算子与局部区域均方差准则结合法改变图像的小波系数,最后对融合后的小波系数进行逆变换,得到重构的目标图像。结果表明,利用该方法不仅降低了融合图像的边缘模糊性,突出了目标色彩,达到良好的视觉效果,而且计算效率高、实时性好,特别有助于伪装目标的检测与识别,具有较好的应用价值。

In order to overcome the defects of fuzzy detection, low recognition rate and poor real-time of traditional fusion methods used in precision-guided weapons systems, a new image fusion algorithm was proposed combining wavelet transform with Canny operator. Firstly, the source image was decomposed into 3 layers in vertical and horizontal directions, which are suitable for image reconstruction; then due to its own characteristics of the different frequency components, an unique fusion rule was used to change wavelet coefficients of images, that is, for the low frequency components, the weighted average fusion algorithm was adopted, and for the high-frequency components, wavelet coefficients were changed using Canny operator and the local area variance criteria method. Finally, images were reconstructed using the inverse wavelet transform for different components. Results show the improved method not only reduces the fuzziness of edge, highlights target color, gets better visual effects, but also makes computational efficiency high, real-time good, particu]ar]y can detect and recognize pretend targets. It has better theoretical research and application value.