Laser & Optoelectronics Progress, Volume. 56, Issue 13, 131001(2019)
Natural Texture Synthesis Algorithm Based on Convolutional Neural Network and Edge Detection
Fig. 1. VGG overall structure
Fig. 2. CNN texture synthesis process
Fig. 3. Edge value of arbitrary point
Fig. 4. Edge extraction by using Kirsch operator. (a) Original image; (b) Kirsch edge extraction effect
Fig. 5. Flow chart of improved algorithm
Fig. 6. Comparison of generated texture images. (a) Input source textures; (b) results of algorithm in Ref. [4]; (c) results of algorithm in Ref. [7]; (d) results of improved algorithm
Get Citation
Copy Citation Text
Dingxiang Zhang, Yongqian Tan. Natural Texture Synthesis Algorithm Based on Convolutional Neural Network and Edge Detection[J]. Laser & Optoelectronics Progress, 2019, 56(13): 131001
Category: Image Processing
Received: Nov. 13, 2018
Accepted: Jan. 22, 2019
Published Online: Jul. 11, 2019
The Author Email: Tan Yongqian (tanyongqian1@163.com)