Research on the Method of Landscape Image Style Transfer based on Semantic Segmentation
Hanmin Ye, Lian Xue, Xiaohui Chen, Wenjie Liu · 2021 IEEE 2nd International Conference on Information Technology, Big Data and Artificial Intelligence (ICIBA) · 2021
The traditional style transfer technique in landscape image has the problem of semantic mismatching in different scenes. In this paper, based on the convolution and landscape style transfer algorithm of neural network, and analyses the semantic segmentation algorithm in theory, based on the normalized statistics of image style transfer algorithm (BN-NST) algorithm was designed on the basis of landscape style transfer algorithm based on semantic segmentation (BS-NST), solve the problem of semantic matching accuracy in different scenarios. In order to verify the algorithm, this paper uses the improved Deeplabv3+ network model and the convolutional neural network algorithm to process the input content image and style image, and the obtained target image has the advantages of authenticity and appreciation of landscape photos.