每一张艺术字的controlnet都不同，需要迭代调整参数。通常canny会用来控制线条，depth控制景深关系，segmentation控制融合度，IP-Adapter控制画面风格。当然这些控制不需要同时使用，而是通过你对图像的一个构思来进行取舍，同时通过权重来影响控制程度。

The controlnet settings are different for each pic and requires iterative adjustment of parameters. Typically canny will be used to control the lines, depth to control the depth of field relationship, segmentation to control the degree of blending, and IP-Adapter to control the style of the image.  these controls don't need to be used at the same time, but rather are trade-offs made by an idea you have for the image, and the degree of control is influenced by the weights.

good luck floks!