Effect Preview

Hello everyone, I am the soul artist AI! Today, I bring you a “beautification + hand-drawing + filter” all-in-one tool: upload a selfie, and poof—automatically transform it into a gentle watercolor sketch. The lines can be thick or thin, shadows adjustable, and it even has built-in skin smoothing and whitening. The entire program is written in Python, using Tkinter for the interface, PIL for filters, and Numpy for efficiency. The structure is layered, clear like a mille-feuille cake. Let’s tackle it layer by layer!

📚 Table of Contents

Project Goals and Knowledge Map
Overall Architecture: Three Main Areas + One Pipeline
Main Window and Global Style
Left Original Image Area: Upload & Preview
Middle Parameter Area: Three Magic Sliders
Right Result Area: Display & Save
Core Hand-drawing Algorithm Breakdown in 8 Steps
Paper Texture Generation Black Technology
Summary & Expandable Easter Eggs

1️⃣ Project Goals and Knowledge Map

Goal	Knowledge Points
Instantly Transform into Hand-drawn Style	PIL Filter Chain, ImageEnhance, ImageFilter
Real-time Preview	Tkinter Scale Variable Binding & Callback
Adjustable Parameters	Sketch Intensity / Detail / Light and Shadow
One-click Save	filedialog & Default Naming
Chinese Interface	ttk.Style + SimHei Font

2️⃣ Overall Architecture: Three Main Areas + One Pipeline

EnhancedHanddrawnConverter
├── create_widgets()          # Three-layer Nine-square Grid Layout
├── upload_image()            # 1. Original Image Entry
├── convert_image()           # 2. Pipeline Control
│   └── apply_enhanced_handdrawn_style()
│       ├── Preprocessing
│       ├── Edge Extraction
│       ├── Face Enhancement
│       ├── Light and Shadow Mask
│       ├── Texture Synthesis
│       └── Final Touch-up
└── save_result()             # 3. Result Output

Reading Order: UI → Parameters → Algorithms → Final Save, smooth like a factory assembly line.

3️⃣ Main Window and Global Style

self.root = root
self.root.title("Enhanced Hand-drawn Style Portrait Converter")
self.root.geometry("1200x750")
self.style = ttk.Style()
self.style.configure(".", font=("SimHei", 10))

Layer 1: The window size is locked at 1200×750 to prevent users from dragging it around and causing layout collapse.
Layer 2: ttk.Style sets the font for all controls to SimHei at once, preventing Chinese characters from garbling, and all subsequent Labels and Buttons inherit this automatically, making it convenient!

4️⃣ Left Original Image Area: Upload & Preview

4.1 Container and Upload Button

left_frame = ttk.LabelFrame(main_frame, text="Original Image", padding=10)
upload_btn = ttk.Button(left_frame, text="Upload Image", command=self.upload_image)
self.original_label = ttk.Label(left_frame, text="Please upload an image")

LabelFrame comes with a title and a recessed line, which is both beautiful and saves code.
padding=10 gives all child controls built-in padding, preventing cramped layout.

4.2 Dynamic Thumbnail

def resize_image(self, image, max_size=400):
    width, height = image.size
    ratio = min(max_size / width, max_size / height)
    new_size = (int(width * ratio), int(height * ratio))
    return image.resize(new_size, Image.LANCZOS)

A general utility method that uniformly resizes any large image to within 400 while maintaining the aspect ratio, using high-quality LANCZOS.
Returns a PIL object for easy reuse later, avoiding repeated decoding.

5️⃣ Middle Parameter Area: Three Magic Sliders

5.1 Layout

param_frame = ttk.LabelFrame(main_frame, text="Style Parameter Adjustment", padding=10)
main_frame.grid_columnconfigure(1, weight=0)   # Fixed Width

The middle column weight=0 ensures that the left and right image areas can stretch, while the middle parameter area maintains a constant width, preventing slider misalignment.

5.2 Three Sliders

self.sketch_scale = Scale(param_frame, from_=0.1, to=1.5, resolution=0.1,
                          command=lambda v: setattr(self, 'sketch_intensity', float(v)))
self.sketch_scale.set(self.sketch_intensity)

setattr(self, 'xxx', float(v)) directly binds instance variables, eliminating the need for an extra IntVar.
resolution=0.1 allows for a step size of 0.1, making dragging smooth and precise.
The three sliders control:

Sketch line intensity (edge thickness)
Detail richness (facial sharpening)
Light and shadow intensity (top soft light)

6️⃣ Right Result Area: Display & Save

right_frame = ttk.LabelFrame(main_frame, text="Conversion Result", padding=10)
self.converted_label = ttk.Label(right_frame, text="The converted image will be displayed here")
save_btn = ttk.Button(right_frame, text="Save Conversion Result", command=self.save_result, state=tk.DISABLED)

The result area is completely symmetrical with the left side, a blessing for perfectionists.
save_btn is disabled by default, only enabled after a successful conversion state=tk.NORMAL, preventing users from clicking on empty space.

7️⃣ Core Hand-drawing Algorithm Breakdown in 8 Steps

Entering the soul segment! A total of 8 layers of filters, stacked like a PS action list.

7.1 Preprocessing: Size & White Background

max_dimension = 1000
if max(width, height) > max_dimension:
    image = image.resize((int(width * ratio), int(height * ratio)), Image.LANCZOS)
white_bg = Image.new('RGB', image.size, (255, 255, 255))

First, resize to prevent a 4000×4000 large image from crashing the memory.
Create a pure white background, ensuring all subsequent blending occurs on white paper to avoid transparent black edges.

7.2 Edge Extraction: Double FIND_EDGES

high_contrast = ImageEnhance.Contrast(image).enhance(1.8)
edges = ImageOps.grayscale(high_contrast).filter(ImageFilter.FIND_EDGES)
edges = ImageEnhance.Contrast(edges).enhance(2.0)

Layer 1: Overall high contrast to capture the main contours.
Layer 2: Further enhance contrast to make the lines darker.

7.3 Detail Edges: DETAIL + FIND_EDGES

detail = image.filter(ImageFilter.DETAIL)
fine_edges = ImageOps.grayscale(detail).filter(ImageFilter.FIND_EDGES)

The DETAIL filter first highlights pore-level details, then FIND_EDGES to prevent “over-smoothing” that blurs facial features.

7.4 Merge & Sketch Base

combined_edges_array = np.minimum(edges_array, fine_edges_array * 0.7)
sketch = Image.blend(gray_contrast, blurred, alpha=self.sketch_intensity)

Numpy vectorization np.minimum is 10 times faster than for-loop, merging thick and thin lines.
blend allows real-time control of sketch intensity with the slider value, making it customizable.

7.5 Face Enhancement (For Female Portraits)

face_region = (width // 4, height // 4, width * 3 // 4, height * 3 // 4)
face_enhanced = Image.fromarray(face).filter(ImageFilter.EDGE_ENHANCE_MORE)
face_enhanced = ImageEnhance.Sharpness(face_enhanced).enhance(self.detail_level)

Simply and brutally treat the center as the face (true CV can replace with dlib / mediapipe), sharpening only the face to ensure eyelashes are distinct.

7.6 Light and Shadow Mask: Top Soft Light

light_mask = Image.new('L', image.size, 200)
draw = ImageDraw.Draw(light_mask)
for i in range(height):
    light_intensity = int(190 + 65 * (1 - i / height * 1.2))
    draw.line([(0, i), (width, i)], fill=light_intensity)

Draw lines row by row to simulate a gradient, brightest at the top and slightly darker at the bottom, instantly creating a “backlit girl by the window” effect.

7.7 Paper Texture Synthesis

texture = self.create_paper_texture(image.size)
final = Image.composite(with_light, white_bg, texture)

Treat the texture as an Alpha mask, where whiter textures are more transparent and darker ones block more, achieving a “pencil on paper” feel.

7.8 Final Touch-up: Contrast + Smoothing + Sharpening

final = ImageEnhance.Contrast(final).enhance(1.2)
final = final.filter(ImageFilter.SMOOTH_MORE)
final = ImageEnhance.Sharpness(final).enhance(1.1)

Enhancing contrast improves overall layering, SMOOTH_MORE removes jagged edges, and finally, light sharpening brings back details, completing the three-step process.

8️⃣ Paper Texture Generation Black Technology

def create_paper_texture(self, size):
    texture = Image.new('L', size, 255)
    draw = ImageDraw.Draw(texture)
    for _ in range(width * height // 200):
        x1 = int(width * (hash(_) % 1000) / 1000)
        y1 = int(height * (hash(_ + 1) % 1000) / 1000)
        opacity = 200 + (hash(_ + 4) % 55)
        draw.line([(x1, y1), (x2, y2)], fill=opacity, width=1)

Using hash() as a pseudo-random number generator, it is cross-platform reproducible and does not require additional imports like random.
Overlaying lines and dots creates a fine texture that resembles sketch paper from a distance, while having a textured noise up close.

9️⃣ Summary & Expandable Easter Eggs

9.1 Knowledge Points Learned

Level	Technical Points
UI	ttk.LabelFrame, Scale, mixed pack/grid layout
Image	PIL filter chain, ImageDraw, Image.composite
Performance	Numpy vectorization, thumbnail downsampling
Interaction	Real-time parameter binding, button state machine
Storage	Automatically generated default file names

9.2 Easter Egg Directions

AI Face Detection: Replace the “assumed face” with precise cropping using face_recognition.
Batch Processing: Add a folder selection to output in bulk *_handdrawn.jpg.
Style Transfer: Integrate Stable Diffusion LoRA for one-click anime, oil painting, cyberpunk.
Web Version: Use Gradio to turn it into an online demo with just three lines of code, making it accessible on mobile.

🏁 Conclusion

In one breath, over 2000 words, we have unfolded the process of “transforming images into watercolor sketches” from UI to algorithms, from variable naming to expandable Easter eggs, layer by layer like peeling an onion. Now you have mastered:

How to use 8 layers of filters to turn photos into hand-drawn images;
How to use three sliders to control effects in real-time;
How to write elegant and maintainable small tools in Python.

Hurry up and run it, give your selfies a one-click “hand-drawn fairy” filter, and amaze everyone!

AI Prompt Words: White Tiger Battle Might

AI Prompt Words: Gothic Mask

AI Prompt Words: Divine Mechanical

AI Prompt Words: Sailor Moon

AI Prompt Words: Six-Winged Angel Gundam

AI Prompt Words: Old Hen

AI Prompt Words: Beauty

AI Prompt Words: Real Person Cosplay

AI Prompt Words: Barbarian Female Warrior

AI Prompt Words: Divine Mechanical

AI Prompt Words: Master Shenlong

AI Prompt Words: Mechanical Playing Cards

AI Prompt Words: Three-Dimensional Miniature Landscape

AI Prompt Words: Gundam

AI Prompt Words: Future Mecha

AI Prompt Words: Female Underworld Fighter

AI Prompt Words: Sun Wukong 3D

AI Prompt Words: Dragon Beast General

AI Prompt Words: Cool Girl Avatar

AI Prompt Words: 9-Grid Emoji Pack

AI Prompt Words: Re-fight

AI Prompt Words: Ink Style Pig

AI Prompt Words: Goddess Nüwa

AI Prompt Words: Guan Yu Hulk

AI Prompt Words: Cool Girl Avatar

AI Prompt Words: Shanhaijing People

AI Prompt Words: Metal Avatar

AI Prompt Words: Classical Beauty

AI Prompt Words: Cartoon Zhang Fei

AI Prompt Words: Beauty Edition

AI Prompt Words: Masked Beauty

Python’s 20-Day Learning Plan

Python’s 7-Day Learning Plan

Python’s Top 10 Basic Libraries

Pygame Module Development for Ball Game

Python Tic-Tac-Toe Mini Game

Python Implementation of Local File Opening Like a Browser

Python Implementation of ID Photo Instant Transformation into “Sakura Cardcaptor”

Python Implementation of Image Color Extractor

Python Implementation of Text-to-Speech Assistant

Python Implementation of Multifunctional 2D Generator

Python Implementation of Foolproof GIF Meme Generator

Python Implementation of Local Camera Viewer

Python Simple Implementation of DeepSeek Q&A Chat

Python Implementation of Video Player

Python Implementation of Simple Notepad

Python Implementation of Multifunctional Application

Python Implementation of Idiom Link Game

Python Implementation of Process Killing System

Python Implementation of Multifunctional Desktop Application

Python Implementation of Volcano API Call to Automatically Generate Text Content

Python Implementation of Doubao AI to Generate Text Stories

Python Implementation of Simple Notepad

Python Implementation of Image Browser Tool Code

Python Implementation of Simple Drawing Tool Code

Python Implementation of To-Do Reminder Tool

Python Implementation of Local Camera Viewer

Python Implementation of Markdown to HTML Tool Code

Python Implementation of 163 Email Push

[Follow for More Updates]

Python: A Very Interesting Game – Source Code for 2048

Python: A Very Interesting Mini Game – Snake

Python: A Super Useful Tool – Word Frequency Statistics Tool

Python: A Practical Tool – Simple Web Scraper Weather Tool

Python: A Super Useful Tool – Timed Task Reminder Tool

Python: “Guess the Number Game Code Analysis”

Python: “Password Generator Code Analysis”

Python: “File Batch Rename Tool Code Analysis”

Python: “Student Grade Management System Code Analysis”

Python: “Stock Data Analysis Tool Code Analysis”

Python + AI Implementation of Intelligent Voice Assistant

Python: “Simple Calculator Code Analysis”

Python + AI Implementation of Online Document Generation Assistant

Python: A Practical Tool – Simple Web Scraper Weather Tool

Python Implementation of Process Killing System

Python Implementation of Link Game Code

Python Implementation of AI Sentiment Analysis Tool

Python Implementation of Local Camera Viewer

Python Implementation of a Simple Real-World Scenario (e.g., Simple Intelligent Q&A System, Text Summary Generation, etc.)

Automatically Transforming Photos into Gentle Watercolor Sketches with Python

Effect Preview

📚 Table of Contents

1️⃣ Project Goals and Knowledge Map

2️⃣ Overall Architecture: Three Main Areas + One Pipeline

3️⃣ Main Window and Global Style

4️⃣ Left Original Image Area: Upload & Preview

4.1 Container and Upload Button

4.2 Dynamic Thumbnail

5️⃣ Middle Parameter Area: Three Magic Sliders

5.1 Layout

5.2 Three Sliders

6️⃣ Right Result Area: Display & Save

7️⃣ Core Hand-drawing Algorithm Breakdown in 8 Steps

7.1 Preprocessing: Size & White Background

7.2 Edge Extraction: Double FIND_EDGES

7.3 Detail Edges: DETAIL + FIND_EDGES

7.4 Merge & Sketch Base

7.5 Face Enhancement (For Female Portraits)

7.6 Light and Shadow Mask: Top Soft Light

7.7 Paper Texture Synthesis

7.8 Final Touch-up: Contrast + Smoothing + Sharpening

8️⃣ Paper Texture Generation Black Technology

9️⃣ Summary & Expandable Easter Eggs

9.1 Knowledge Points Learned

9.2 Easter Egg Directions

🏁 Conclusion

Leave a Comment Cancel reply

Effect Preview

📚 Table of Contents

1️⃣ Project Goals and Knowledge Map

2️⃣ Overall Architecture: Three Main Areas + One Pipeline

3️⃣ Main Window and Global Style

4️⃣ Left Original Image Area: Upload & Preview

4.1 Container and Upload Button

4.2 Dynamic Thumbnail

5️⃣ Middle Parameter Area: Three Magic Sliders

5.1 Layout

5.2 Three Sliders

6️⃣ Right Result Area: Display & Save

7️⃣ Core Hand-drawing Algorithm Breakdown in 8 Steps

7.1 Preprocessing: Size & White Background

7.2 Edge Extraction: Double FIND_EDGES

7.3 Detail Edges: DETAIL + FIND_EDGES

7.4 Merge & Sketch Base

7.5 Face Enhancement (For Female Portraits)

7.6 Light and Shadow Mask: Top Soft Light

7.7 Paper Texture Synthesis

7.8 Final Touch-up: Contrast + Smoothing + Sharpening

8️⃣ Paper Texture Generation Black Technology

9️⃣ Summary & Expandable Easter Eggs

9.1 Knowledge Points Learned

9.2 Easter Egg Directions

🏁 Conclusion

Related posts

Leave a Comment Cancel reply