Translation line UAV inspection Image quality assessment based on Salient edge Blur Metric

Autor: Liu Guangxiu, Xiaobin Sun, Yue Liu, Wanguo Wang, Shiyou Mu
Rok vydání: 2019
Předmět:
Zdroj: 2019 IEEE 4th Advanced Information Technology, Electronic and Automation Control Conference (IAEAC).
DOI: 10.1109/iaeac47372.2019.8997754
Popis: The recent development of UAV in translation line inspection leads to more efficient inspection than human. The image quality of UAV is an import rule to judge the UAV inspection effectiveness. A global and local features fusion based method is presented for UAV image quality evaluation in this paper. For the evaluation of overall quality of image, it combines sharpness index in frequency domain and the distribution density and orientation features of edge lines in space domain to judge the existence of the camera shake blur. And the average width of salient edges is analyzed to determine the blur degree of local details in image. Finally, this evaluation method is verified effectively through many inspection images. It shows its effectiveness and robustness under various complex background and different contents of the image, and the evaluation result is consistent with 5 standards of subjective evaluation. So the method will be suitable to the intelligent recognition and analysis.
Databáze: OpenAIRE