Catenary image enhancement method based on Curvelet transform with adaptive enhancement function

Abstract

In the process of catenary failure diagnosis system based on image processing technique, some catenary images present low contrast, which need to be enhanced. Curvelet transform has the high directional sensitivity and anisotropy, which is suitable for image enhancement because of its optimal sparse representation of image with rich details and edges. First, the catenary image is decomposed by Curvelet transform to get its high and low frequency coefficients, then adjust the high frequency coefficients using the enhancement function. Afterwards, combine the high frequency coefficients and low frequency coefficients by the inverse Curvelet transform, and thus to get the enhanced catenary image. In this paper, Curvelet transform is compared with the traditional enhancement methods. The experimental results show that the proposed method can effectively enhance the low contrast catenary images, the catenary insulator, arm, hanger, pillar and locator part become visible, the details become more obvious. Moreover, as for the online application of catenary failure diagnosis system, efficiency is another important consideration. The experimental results also show that the cost time of catenary image enhancement is within a few tens of seconds, which meets the requirements of catenary failure diagnosis system.

Multi-Focus Image Fusion with Multi-Scale Transform Optimized by Metaheuristic Algorithms

“…The coarsest layer is the low-frequency sub image information and represents the main information of the image. The finest layer contains the highfrequency sub-images information and noise which reflects the details components and detail layer contains the strong edge details information of the source image from different directions and scales [32]. Commonly for the fusion, while an average or weighted average rule is used in the low-frequency sub image, the maximum absolute rule is often used in the high frequency sub image.…”