Generative AI tools have brought unprecedented convenience to image creation. Whether it’s artwork, design concepts, or personal projects, you can generate high-quality images just by entering text descriptions (Prompts). In this article, we’ll teach you how to create images using generative AI tools, using Copilot as an example, and explain each step and technique in detail.

---

What Are Generative AI Image Tools?

Generative AI image tools are technologies that create images through artificial intelligence. By simply entering descriptive text (Prompts), AI can quickly generate realistic or artistic images. Tools like Copilot, MidJourney, DALL-E, and Stable Diffusion are widely used in design, advertising, artistic creation, and other fields.

---

The Importance and Basic Structure of Prompts

Prompts are instructions that help AI understand your needs. They define the subject, style, details, and mood of the image. Writing excellent prompts is key to generating high-quality images.

Basic Prompt Structure:

• Main Description: What you want to generate (e.g., a girl riding a unicorn).
• Details: Including colors, lighting, background, etc. (e.g., soft morning light, forest background).
• Style and Mood: Specify artistic style or emotion (e.g., romantic oil painting style, peaceful atmosphere).
• Excluded Elements (optional): Specify elements you don’t want to appear in the image.

---

10 Key Elements for Image Generation

When writing prompts, consider these key elements to make your descriptions clearer:

Element	Description	Example
1. Subject Description	Describe the main subject or scene you want to generate	A girl riding a unicorn
2. Style & Art Type	Specify artistic style or type, such as realistic, anime, surrealism, etc.	Vintage city night scene in oil painting style
3. Color & Lighting	Define image tone and lighting effects, like cool tones, warm tones, soft light, etc.	Warm tones with soft morning light
4. Background & Environment	Describe the background or environment of the main subject	Interior of a futuristic space station
5. Action or Movement	Describe the subject’s action or state, like running, jumping, standing still	A fox running joyfully
6. Clothing & Details	Describe character clothing, accessories, or other details	A knight wearing medieval armor
7. Time & Era	Set the time or historical era of the image	Medieval castle at dusk
8. Perspective & Composition	Specify desired image perspective and layout, like bird’s-eye view, low angle, close-up, panorama	Bird’s-eye view of a futuristic city street
9. Emotion & Atmosphere	Specify the emotions or atmosphere to convey, such as romantic, horror, peaceful, etc.	Dreamy and peaceful lake
10. Excluded Elements	Define objects or elements you don’t want to appear	No humans

---

Practical Guide Using Copilot

Step 1: Access Copilot Chat

• Go to Copilot Chat. (Remember to log in if you haven’t already)

Step 2: Write Your Prompt

Use the 10 elements above to design your prompt and enter it in the text box.

Example Prompt:

A white unicorn standing by a clear lake in a forest, oil painting style, with soft morning light, misty forest in the background, romantic and peaceful atmosphere.

Step 3: Generate the Image

• Press the “Enter” key or click the [Submit message] button on the right.

• Review the generated result. If you’re not satisfied, adjust the prompt by adding more details.

Step 4: Download the Image

• Download the image.

---

Common Issues and Best Practices

Issue 1: Image Generation Results Don’t Meet Expectations?

Reasons:

• Prompt is too brief, lacking descriptive details, making it difficult for AI to understand your requirements accurately.
• Missing supporting information like background, lighting, or style, which may result in randomly generated images.

Example 1:

• Poor Prompt:

Draw a unicorn

Issues Identified:

1. Lacks detail about the subject, such as color, size, or action.
2. No defined background (is it in a forest, desert, or space?).
3. No specified artistic style (realistic, cartoon, or oil painting style?).

• Improved Prompt:

A white unicorn standing by a clear lake in a forest, oil painting style, with soft morning light, misty forest in the background, romantic and peaceful atmosphere.

Why is the Improved Version Better?

1. Provides specific details about the subject (white, standing).
2. Clarifies the background (forest, lakeside, morning light).
3. Includes style and atmosphere descriptions (oil painting style, romantic atmosphere).

---

Issue 2: How to Make Images More Artistic?

Reason:

• The prompt lacks descriptions of style, color, and mood, which may result in images that appear ordinary or lack artistic appeal.

Example 2:

• Poor Prompt:

Draw a village

Issues Identified:

1. Description is too vague, can’t determine the specific appearance of the village (modern, ancient, or fantasy?).
2. No specified perspective (bird’s-eye view, close-up, or distant view).
3. No emphasis on colors or atmosphere (warm, cool, or romantic?).

• Improved Prompt:

A bird's-eye view of a medieval European village with stone buildings and thatched roofs, sunlight filling the streets, warm color tones, vintage illustration style, peaceful and serene atmosphere.

Why is the Improved Version Better?

1. Defines the village’s era and architectural features (medieval, stone buildings).
2. Specifies the perspective (bird’s-eye view).
3. Includes color and style elements (warm tones, vintage illustration style).

---

Issue 3: How to Exclude Unwanted Elements?

Reason:

• AI might automatically generate elements you don’t want to see, such as humans, text, or background objects.

Solution:

• Add “excluded elements” in your prompt to clearly tell the AI what content to avoid generating.

Example:

• Poor Prompt:

Draw a lion on the savanna.

Issues Identified:

1. AI might generate additional animals or human characters on the savanna, distracting from the main subject.
2. No explicit exclusion of unwanted elements, such as buildings or text in the background.

• Improved Prompt:

A lion with golden mane standing on a vast savanna, background with only natural landscape, no humans or other animals.

Why is the Improved Version Better?

1. Clearly specifies that the background should only contain natural landscape.
2. Includes the condition “no humans or other animals,” helping the AI generate images that more accurately meet the requirements.

---

Issue 4: Why Aren’t the Image Details Rich Enough?

Reason:

• The prompt lacks detailed descriptions, failing to guide the AI in adding richer content.

Example:

• Poor Prompt:

Draw a futuristic city

Issues Identified:

1. No description of specific city features (such as building shapes, transportation).
2. Lacks description of lighting and atmosphere.

• Improved Prompt:

A bird's-eye view of a futuristic city with transparent glass structures, flying cars moving through the air, a massive energy tower in the city center, neon lighting effects, starry sky background, and a high-tech atmosphere.

Why is the Improved Version Better?

1. Provides specific architectural features (transparent glass structures).
2. Includes scene details (flying cars, massive energy tower).
3. Describes lighting effects (neon lighting, starry sky background).

---

Conclusion

Writing high-quality prompts is the key skill for successful use of generative AI tools like Copilot. By crafting detailed, clear descriptions and avoiding vague or contradictory content, you can help AI generate images that better meet your requirements. Whether you’re a beginner or an advanced user, following the methods in this tutorial will help you create stunning works! Start experimenting now and explore the unlimited creative possibilities of generative AI!