基于同态滤波和直方图均衡化的图像增强算法

Image enhancement algorithm based on homomorphic filtering and histogram equalization

基于图像的目标识别是构建智能交通的基础,对提高智能驾驶的安全性具有重要意义。考虑到实际环境下图像质量由于光照不均匀影响而出现不同程度的降低,提出了一种结合同态滤波和直方图均衡化的图像增强算法。首先基于频域增强处理的同态滤波,研究了一种单参数滤波算法,减少调节参数,校正光照不均影响,提高图像清晰度。然后应用基于线性插值的对比度受限的自适应直方图均衡化算法对图像HSV色彩空间中的透明度分量进行非线性拉伸,提高图像对比度。对该算法、传统高斯型同态滤波算法、对比度受限的自适应直方图均衡化算法进行了仿真对比试验,实验结果表明,该算法能有效改善光照不均匀对图像的影响,增强图像细节、提高图像对比度和清晰度,可提高图像识别的准确度和环境适应性。

Image based target recognition is the basis of intelligent transportation, which is of great significance to improve the safety of intelligent driving. Considering that the image quality in the actual environment is reduced to different degrees due to the influence of uneven illumination, an image enhancement algorithm combining homomorphic filtering and histogram equalization is proposed. First, based on the homomorphic filtering of frequency domain enhancement processing, a single-parameter filtering algorithm is proposed to reduce the adjustment parameters, correct the influence of uneven illumination, and improve image clarity. Then, the contrast limited adaptive histogram equalization algorithm that based on linear interpolation is applied to nonlinearly stretch the transparency components in the HSV color space of the image to enhance the contrast of the image. This algorithm, the traditional Butterworth homomorphic filtering algorithm, and the contrast-limited histogram equalization algorithm are simulated and compared. The experimental results show that the algorithm can effectively improve the impact of uneven illumination on the image, enhance image details, and improve image contrast and clarity, and improve the accuracy of image recognition and environmental adaptability.