How to use midjourney to draw pictures Midjourney is a mainstream foreign AI drawing tool. It can not only randomly generate pictures, but also provides a way to use reference pictures/cushion pictures to make drawings more creative. When using Midjourney, there are two main ways to use reference images/mat images: use reference images + keywords to describe the image and use the blend command to fuse different elements to generate a set of fused images.
1. Use reference pictures + keywords to describe the pictures
·Upload reference pictures: In Midjourney, click the plus sign to upload the pictures you need for reference. Midjourney Bot will generate an image based on this image.
·Naming: Midjourney Bot's own system can only recognize links. You need to enter a name in the dialog box and press Enter. After naming, click on the image, open it in a browser, and copy the image URL.
Click on the image to open it in your browser
Copy image URL
·Descriptor: Paste the copied image link behind the imgine prompt. Note that the link suffix must be .png or .jpg.
Then you need to leave another space, enter the description, and press the Enter key. The newly generated image will refer to the original image.
Paste the copied image link behind the imgine prompt, leave a space, enter a description, and press Enter. Use the --iw command to control the similarity between the new image and the reference image.
2. Midjourney uses the blend command. This command can fuse several pictures well and generate a set of fused pictures. Select the blend command: Enter / in the plus column and select the blend command.
Add pictures: You can add up to five pictures for fusion. After the addition is completed, press the Enter key to generate a set of fusion pictures, which is suitable for merging different elements or rendering styles of similar subjects.
3. Select the describe command
Drag in pictures
Press Enter to automatically identify the picture description and related parameters.
You can translate these 4 types of analysis results, select the description you want most, and press the corresponding number to generate a picture.
After pressing submit, you can wait for the image to be generated. If you are not satisfied with the result, you can copy the text and make slight changes before generating it again.