Road object extraction method based on saliency in night vision infrared image

摘要

First, a saliency region is roughly extracted based on local contrast using a GBVS model (an image saliency analysis algorithm based on graph theory), and a saliency of a night vision infrared image is obtained. Re-extraction of the saliency region is performed on the saliency image CC based on the overall features using the step S1 for acquiring the image CC and a method based on the spectral scale space of the super complex frequency region, and night vision By using step S2 for acquiring the saliency image ZZ of the infrared image and the saliency extraction method based on the fusion of the whole and local information, the whole and the local information are fused to the saliency image ZZ. Step S3 for obtaining the saliency image AA. [Selection] Figure 1