Skip to content

00_Overview.md

Latest commit

 

History

History
305 lines (228 loc) · 11.7 KB

00_Overview.md

File metadata and controls

305 lines (228 loc) · 11.7 KB

OpenCV / Computer Vision Learning Guide

Introduction

This folder contains systematic learning materials for computer vision using OpenCV. You can learn step-by-step from the basics of image processing to deep learning-based object detection.

Target Audience: Developers with Python basics, computer vision beginners, those preparing for video processing projects

Learning Roadmap

[Basics]                 [Intermediate]             [Advanced]
  │                         │                         │
  ▼                         ▼                         ▼
Setup (01) ─────────▶ Filtering (05) ──────▶ Feature Detection (13)
  │                         │                         │
  ▼                         ▼                         ▼
Image Basics (02) ──▶ Morphology (06) ─────▶ Feature Matching (14)
  │                         │                         │
  ▼                         ▼                         ▼
Color Spaces (03) ──▶ Thresholding (07) ───▶ Object Detection (15)
  │                         │                         │
  ▼                         ▼                         ▼
Geometric Transforms (04) ─▶ Edge Detection (08) ──▶ Face Detection (16)
                            │                         │
                            ▼                         ▼
                     Contours (09) ─────────▶ Video Processing (17)
                            │                         │
                            ▼                         ▼
                     Shape Analysis (10) ───▶ Calibration (18)
                            │                         │
                     ┌──────┴──────┐                  ▼
                     ▼             ▼           DNN Module (19)
               Hough Transform (11)  Histogram (12)        │
                                                      ▼
                                              Practical Projects (20)

Prerequisites

Required

  • Python basics (variables, control flow, functions, classes)
  • NumPy basics (ndarray, indexing, slicing, broadcasting)
  • File I/O, exception handling

Recommended

  • Linear algebra basics (matrix operations)
  • Probability/statistics basics
  • Machine learning concepts (classification, training)

File List

Basics (01-04)

File Difficulty Key Topics
01_Environment_Setup.md ⭐ OpenCV installation, opencv-python vs contrib, version check
02_Image_Basics.md ⭐ imread, imshow, imwrite, pixel access, ROI
03_Color_Spaces.md ⭐⭐ BGR/RGB, HSV, LAB, cvtColor, channel separation
04_Geometric_Transforms.md ⭐⭐ resize, rotate, flip, warpAffine, warpPerspective

Image Processing (05-08)

File Difficulty Key Topics
05_Image_Filtering.md ⭐⭐ blur, GaussianBlur, medianBlur, bilateralFilter
06_Morphology.md ⭐⭐ erode, dilate, opening, closing, gradient
07_Thresholding.md ⭐⭐ threshold, OTSU, adaptiveThreshold
08_Edge_Detection.md ⭐⭐⭐ Sobel, Scharr, Laplacian, Canny

Object Analysis (09-12)

File Difficulty Key Topics
09_Contours.md ⭐⭐⭐ findContours, drawContours, hierarchy, approxPolyDP
10_Shape_Analysis.md ⭐⭐⭐ moments, boundingRect, convexHull, matchShapes
11_Hough_Transform.md ⭐⭐⭐ HoughLines, HoughLinesP, HoughCircles
12_Histogram_Analysis.md ⭐⭐ calcHist, equalizeHist, CLAHE, backprojection

Features and Detection (13-15)

File Difficulty Key Topics
13_Feature_Detection.md ⭐⭐⭐ Harris, FAST, SIFT, ORB, keypoints/descriptors
14_Feature_Matching.md ⭐⭐⭐ BFMatcher, FLANN, ratio test, homography
15_Object_Detection_Basics.md ⭐⭐⭐ template matching, Haar cascade, HOG+SVM

Advanced Topics (16-18)

File Difficulty Key Topics
16_Face_Detection.md ⭐⭐⭐⭐ Haar/dlib face detection, landmarks, LBPH, face_recognition
17_Video_Processing.md ⭐⭐⭐ VideoCapture, VideoWriter, background subtraction, optical flow
18_Camera_Calibration.md ⭐⭐⭐⭐ camera matrix, distortion correction, chessboard calibration

DNN and Practice (19-20)

File Difficulty Key Topics
19_DNN_Module.md ⭐⭐⭐⭐ cv2.dnn, readNet, blobFromImage, YOLO, SSD
20_Practical_Projects.md ⭐⭐⭐⭐ document scanner, lane detection, AR marker, face filter

3D Vision (21-23)

File Difficulty Key Topics
21_3D_Vision_Basics.md ⭐⭐⭐ stereo vision, depth map, point cloud, 3D reconstruction
22_Depth_Estimation.md ⭐⭐⭐⭐ monocular depth estimation, MiDaS, DPT, Structure from Motion
23_SLAM_Introduction.md ⭐⭐⭐⭐ Visual SLAM, ORB-SLAM, LiDAR SLAM, Loop Closure

Segmentation (25-27)

File Difficulty Key Topics
25_Semantic_Segmentation.md ⭐⭐⭐⭐ FCN, U-Net, DeepLab, pixel-wise classification
26_Instance_Segmentation.md ⭐⭐⭐⭐ Mask R-CNN, YOLACT, instance-level masks
27_Panoptic_Segmentation.md ⭐⭐⭐⭐ Panoptic FPN, stuff + things, unified segmentation

Neural 3D and Novel View Synthesis (28-29)

File Difficulty Key Topics
28_Neural_Radiance_Fields.md ⭐⭐⭐⭐⭐ NeRF, volumetric rendering, positional encoding
29_3D_Gaussian_Splatting.md ⭐⭐⭐⭐⭐ 3DGS, differentiable rasterization, real-time rendering

Video and Data (30-32)

File Difficulty Key Topics
30_Video_Understanding.md ⭐⭐⭐⭐ C3D, I3D, SlowFast, Video Transformers, action recognition
31_Optical_Flow.md ⭐⭐⭐⭐ Lucas-Kanade, Farneback, RAFT, FlowNet
32_Synthetic_Data_Generation.md ⭐⭐⭐⭐ Domain randomization, sim-to-real, data augmentation

Recommended Learning Sequence

Quick Start (1 week)

01 → 02 → 03 → 05 → 07 → 08 → 09

Basic Completion (2-3 weeks)

01 → 02 → 03 → 04 → 05 → 06 → 07 → 08 → 09 → 10 → 11 → 12

Intermediate Completion (4-5 weeks)

Basic Completion + 13 → 14 → 15 → 16 → 17

Full Mastery (6-8 weeks)

Intermediate Completion + 18 → 19 → 20

3D Vision Deep Dive (2 weeks)

Full Mastery + 21 → 22 → 23

Segmentation and Modern Topics (3-4 weeks)

3D Vision + 25 → 26 → 27 → 28 → 29 → 30 → 31 → 32

Practice Environment

Installation

# Basic installation (most features)
pip install opencv-python numpy matplotlib

# Extended installation (additional features like SIFT, SURF)
pip install opencv-contrib-python

# For face recognition
pip install dlib face_recognition

# Version check
python -c "import cv2; print(cv2.__version__)"

Recommended Environment

- Python 3.8 or higher
- OpenCV 4.x
- IDE: VSCode, PyCharm, Jupyter Notebook
- OS: Windows, macOS, Linux all supported

Example Project Structure

my_cv_project/
├── images/           # Input images
├── output/           # Output results
├── models/           # Trained models (Haar, DNN, etc.)
├── src/              # Source code
│   ├── preprocessing.py
│   ├── detection.py
│   └── utils.py
├── notebooks/        # Jupyter experiments
└── requirements.txt

OpenCV Quick Function Reference

Image I/O

Function Description Example
cv2.imread() Read image img = cv2.imread('image.jpg')
cv2.imshow() Display image cv2.imshow('Window', img)
cv2.imwrite() Save image cv2.imwrite('out.jpg', img)

Color Conversion

Function Description Example
cv2.cvtColor() Color space conversion gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.split() Split channels b, g, r = cv2.split(img)
cv2.merge() Merge channels img = cv2.merge([b, g, r])

Geometric Transforms

Function Description Example
cv2.resize() Resize resized = cv2.resize(img, (w, h))
cv2.rotate() Rotate rotated = cv2.rotate(img, cv2.ROTATE_90_CLOCKWISE)
cv2.warpAffine() Affine transform dst = cv2.warpAffine(img, M, (w, h))

Filtering

Function Description Example
cv2.GaussianBlur() Gaussian blur blur = cv2.GaussianBlur(img, (5, 5), 0)
cv2.Canny() Canny edge detection edges = cv2.Canny(img, 100, 200)
cv2.threshold() Thresholding _, binary = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY)

Contours/Shapes

Function Description Example
cv2.findContours() Find contours contours, _ = cv2.findContours(binary, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cv2.drawContours() Draw contours cv2.drawContours(img, contours, -1, (0, 255, 0), 2)
cv2.boundingRect() Bounding rectangle x, y, w, h = cv2.boundingRect(contour)

Related Resources

Links to Other Folders

Folder Related Content
Python/ Advanced Python syntax, testing, packaging
Algorithm/ Algorithms for image processing (graphs, DP)
Linux/ Development environment, file handling

External Resources

Learning Tips

  1. Practice-oriented: Execute code directly and check results
  2. Visualization: Visualize intermediate results with matplotlib
  3. Parameter tuning: Real-time adjustment with trackbars (createTrackbar)
  4. Step-by-step processing: Break down complex tasks into pipelines
  5. Debugging: Check images at each step with imshow

Interview Preparation Key Topics

Topic Key Questions
Color spaces RGB vs HSV - When to use HSV?
Filtering Gaussian vs Bilateral - What's the difference?
Thresholding What is the principle of Otsu's method?
Edge detection What are the steps of Canny algorithm?
Features SIFT vs ORB - Compare pros and cons
Object detection Haar cascade vs HOG+SVM differences
DNN YOLO vs SSD - Speed and accuracy trade-offs