In this section we will continue to delve deeper into ControlNet's line constraints capabilities, including Linework Mode, Graffiti Mode, and Line Mode. These modes will help us better understand and apply the role of line constraints in image generation.
I. Lineart lineart mode: more natural edge extraction
In the previous content, we discussed the limitations of hard and soft edge modes. Hard edge mode tends to cause double line problems, while soft edge mode has weak constraints. To address these issues, ControlNet version 1.1 introduces the Line Drawing mode, which extracts lines more naturally without double edges and intelligently adjusts the line thickness based on the thickness of the edge.
1、Application cases of line drawing mode
We show how the line drawing mode can be effective in extracting a single line instead of a double edge through the case of an anime style cat. The emergence of this mode makes us prefer to use the line drawing mode over the hard or soft edge modes in most cases.
2. Differences in preprocessors
There are five preprocessors in line drawing mode, and the difference between them is mainly in the level of detail in the lines. For example, anime line drawing extraction pays more attention to the preservation of facial features, while denoising mode reduces the extraction of details and makes the lines more concise. These pre-processors can be chosen flexibly according to actual needs.
Second, Scribble graffiti mode: free play line constraints
Doodle mode is very different from line drawing mode, which focuses on free play and brainstorming, and extracts more rough lines, giving the AI more room for creativity. In this mode, the AI can generate interesting images even if the lines are of low quality, making it particularly suitable for applications such as children's education and assisted composition.
Doodle mode generates imaginative images. This mode helps children transform simple doodles into beautiful drawings, and also assists designers with composition and product fine-tuning.
III. MLSD straight-line model: focusing on straight-line extraction
Straight line mode focuses on extracting straight lines in an image, ignoring curves and circular lines, making it uniquely useful in areas such as architectural design. This mode can help us quickly extract architectural frames for interior design.
The straight line mode can effectively extract the straight line structure of the interior space and redesign it on that basis. This mode performs well in processing images with obvious straight line structure.
IV. Conclusion
In this section, we have learned about the different modes of line constraints in ControlNet and their applications. Hard Edge, Soft Edge, Line Drawing, Graffiti and Straight Line modes each have their own characteristics, and we need to choose and apply these modes flexibly according to the specific needs of the project.
I hope this article will help you get a better grasp of ControlNet's line constraints feature.
