https://github.com/merteldem1r/opencv-core

OpenCV course to manipulate images and videos, and detect objects and faces, among other exciting topics.
https://github.com/merteldem1r/opencv-core

computer-vision opencv python

Last synced: about 2 months ago
JSON representation

OpenCV course to manipulate images and videos, and detect objects and faces, among other exciting topics.

• Host: GitHub
• URL: https://github.com/merteldem1r/opencv-core
• Owner: merteldem1r
• Created: 2025-07-09T21:30:25.000Z (12 months ago)
• Default Branch: main
• Last Pushed: 2025-07-15T16:02:14.000Z (12 months ago)
• Last Synced: 2025-07-16T11:27:54.235Z (12 months ago)
• Topics: computer-vision, opencv, python
• Language: Jupyter Notebook
• Homepage:
• Size: 21.3 MB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
## OpenCV Core

Core concepts & implementations of OpenCV library to manipulate images and videos, and detect objects and faces, among other topics.

- **Course**: [OpenCV University](https://opencv.org/university/)

- **Implemented by**: [Mert Eldemir](https://github.com/merteldem1r)

**NOTE**: To manipulate code and see the results first download the respective images and documents. On each module folder run the `download_assets.py` script

### Module 1: Getting Started With Images

- Reading & Display images using **opencv** and **matplotlib**

- Representing images as **numpy arrays** & **RGB** & **HSV** format images

- **cv2**: `imread()` `imwrite()` `cvtColor()` `split()` `merge()`

  - File: [image-handling.ipynb](01-Image-Handling/notebooks/image_handling.ipynb)

### Module 2: Basic Image Manipulation

- **Accessing** & **Modifying** image pixels

- **Crop** & **Resize** & **Flipping** images

- **cv2**: `resize()` `flip()`

  - File: [image_manipulation.ipynb](02-Image-Manipulation/notebooks/image_manipulation.ipynb)

### Module 3: Image Annotation

- **Annotating** images

- Drawing **line**, **circle**, **rectangle** & Putting **text** on images

- **cv2**: `line()` `circle()` `rectangle()` `putText()`

  - File: [image_annotation.ipynb](03-Image-Annotation/notebooks/image_annotation.ipynb)

### Module 4: Image Enhancement

- **Brightness** & **Contrast** & **Image Thresholding**

- Difference between **Global** and **Adaptive thresholding**

- Thresholding Types: **THRESH_BINARY**, **THRESH_BINARY_INV**, **THRESH_TRUNC** etc.

- Bitwise operations: **AND**, **OR**, **XOR**, **NOT**

- Applying **Mask** on the background

- Creating and type casting **numpy arrays**

- **cv2**: `add()` `subtract()` `multiply()` `threshold()` `adaptiveThreshold()` `bitwise_and()` `bitwise_or()` `bitwise_xor()` `bitwise_not()`

  - File: [image_enhancement.ipynb](04-Image-Enhancement/notebooks/image_enhancement.ipynb)

### Module 5: Accessing and Writing to Camera

- **Accessing** to camera and **Writing video** as **mp4**

- **cv2**: `VideoCapture()` `VideoWriter()`

  - File: [camera.py](05-Access-Write-Camera/camera.py)

### Module 6: Image Filtering and Edge Detection

- Opening camera via **VideoCapture** and using keyboard to apply different **filters** and **detection functions**

- **Edge** & **Corner** & **Filter (Kernel)** term explanations

- **Gaussian Blur** & **Edge Detection** & **Corner Detection** (Shi-Tomasi)

- **cv2**: `blur()` `GaussianBlur()` `Canny()` `goodFeaturesToTrack()`

  - File: [detection.py](06-Image-Filtering-and-Edge-Detection/detection.py)

### Module 7: Image Features and Alignment

- Goal is **warping** scanned image to make it look as original reference image (basically align them)

- Using **Feature-based matching** (ORB) and **Homography** (a 3×3 transformation matrix that maps points from one perspective to another)

- **KeyPoint** & **Descriptor** & **Matching**

- **Use cases**: Document alignment & scanning, Panorama stitching, Augmented Reality (placing virtual objects correctly) etc.

- **cv2**: `ORB_create()` `orb.detectAndCompute()` `drawKeypointss()` `DescriptorMatcher_create()` `drawMatches()` `findHomography()` `warpPerspective()`

  - File: [image_features_alignment.ipynb](07-Image-Features-and-Alignment/notebooks/image_features_alignment.ipynb)

### Module 8: Panorama

- Image allignment and creating **panorama** image

- **cv2**: `Stitcher_create()` `stitcher.stitch`

  - File: [panorama.ipynb](08-Panorama/notebooks/panorama.ipynb)

### Module 9: HDR

- Idea of **HDR** Images

- **MTB - Median Threshold Bitmap** (compute the median pixel intensity)

- **cv2**: `createAlignMTB()` `alignMTB.process()` `createCalibrateDebevec()` `calibrate_debevec.process()` `createMergeDebevec()` `createTonemapDrago()` `createTonemapReinhard()` `createTonemapMantiuk()`

  - File: [hdr.ipynb](09-HDR/notebooks/hdr.ipynb)

### Module 10: Object Tracking

- **Object Detection** vs **Object Tracking**

- **Object Tracking** steps & differences on **Deep learning-based models**

- Tracking libraries like **GOTURN**, **MIL**, **Nano**, **Vit**, **mean shift** etc.

- **Appearance Model** & **Motion Model** terms

- **cv2**: `TrackerMIL()` `TrackerGOTURN()` `TrackerNano()` `dnn.readNet()` `TrackerVit()` `getTickCount()` `getTickFrequency()`

  - File: [tracker.py](10-Object-Tracking/tracker.py)

  - Notebook File: [hdr.ipynb](10-Object-Tracking/notebooks/object_tracking.ipynb)

### Module 11: Face Detection

- **Real-time face detection** using a **pre-trained deep neural network (DNN)**

- Loads **Caffe model** via **deploy.prototxt** and **.caffemodel** for inference

- Converts input frame to **blob** and feeds into the network

- Performs forward pass to obtain **detection results**

- Iterates over **4D Numpy output tensor (detections) to extract face boxes**

- Draws **bounding boxes**, **confidence scores**, and **inference time**

- **cv2**: `getTextSize()` `dnn.readNetFromCaffe()` `dnn.blobFromImage()` `net.setInput()` `net.forward()` `net.getPerfProfile()`

  - File: [faceDetection.py](11-Face-Detection/faceDetection.py)

### Module 12: TensorFlow Object Detection

- **TensorFlow** Object Detection **manual Setup**

- Downloading and extracting pre-trained `ssd_mobilenet_v2_coco_2018_03_29.tar.g ` (trained on **COCO dataset**)

- Using inference file `frozen_inference_graph.p ` with external `tf_text_graph_ssd.py` script to get OpenCV usable `ssd_mobilenet_v2_coco_2018_03_29.pbtx` file

- Using `coco_class_labels.txt`to clasify found objects

- Performing **inferenece** using a **OpenCV DNN**

- Loading the model and input image into memory, detecting objects using a forward pass through the network

- Display the detected objects with **bounding boxes** and **class labels**

- **cv2**: `dnn.readNetFromTensorflow()`

  - File: [ts_object_detection.ipynb](12-TensorFlow-Object-Detection/notebooks/ts_object_detection.ipynb)

### Module 13: Pose Estimation using OpenPose

- Using Caffe model trained on the **Multi-Person Image Dataset (MPI)** to demonstrate **human pose estimation for a single person**.

- Loading **Caffe model**:

  - **Architecture**: `pose_deploy_linevec_faster_4_stages.prototxt`

  - **Weights**: `pose_iter_160000.caffemodel`

- Converting to blog for wrapping image into format that DNN expects

- Visualizing **probability map (heatmap)** of found body parts

- Displaying **Body Points** & **Skeleton** on the input image

  - File: [pose_estimation.ipynb](13-Pose-Estimation-OpenPose/notebooks/pose_estimation.ipynb)