Skip to main content

From brush painting to high relief.

Fu, Y., 2019. From brush painting to high relief. Doctoral Thesis (Doctoral). Bournemouth University.

Full text available as:

FU, Yunfei_Ph.D._2019.pdf



As an artistic form, relief is described as a hybrid between 2D painting and 3D sculpture. A novel approach for generating a texture-mapped high relief model from a single brush painting is presented in this work. The aim of this work is to extract the brush strokes from a painting and generate the individual corresponding relief proxies rather than recovering the exact depth map from the painting, which is a tricky computer vision problem, requiring assumptions that are rarely satisfied. The relief proxies of brush strokes are then combined together to form a 2.5D high-relief model. To extract brush strokes from 2D paintings, this work applies layer decomposition and stroke segmentation by imposing boundary constraints. The segmented brush strokes preserve the style of the input painting. By inflation and a displacement map of each brush stroke, the features of brush strokes are preserved by the resultant high relief model of the painting. As the adjacent brush strokes may share similar colours in some brush paintings, the layer decomposition method does not work well. To amend this issue, this work also proposed a deep learning based method for brush stroke extraction. This work demonstrates that it is able to produce convincing high-reliefs from a variety of paintings (with humans, animals, flowers, etc.). As a secondary application, this work shows how the proposed brush stroke extraction algorithm could be used for image editing. As a result, the proposed brush stroke extraction algorithm is specifically geared towards paintings with each brush stroke drawn very purposefully, such as Chinese paintings, Rosemailing paintings, etc.

Item Type:Thesis (Doctoral)
Additional Information:If you feel that this work infringes your copyright please contact the BURO Manager.
Uncontrolled Keywords:high relief
Group:Faculty of Media & Communication
ID Code:32757
Deposited By: Symplectic RT2
Deposited On:13 Sep 2019 09:37
Last Modified:14 Mar 2022 14:17


Downloads per month over past year

More statistics for this item...
Repository Staff Only -