tune up

Author: anhphan2705

.gitignore

@@ -159,7 +159,4 @@ cython_debug/
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
 
-data/
-output/
-background_subtractor.py
-static_differences.py
+data/

README.md

@@ -1,22 +1,24 @@
-# Video Object Detection and Tracking
+# Object Detection Camera Feed 
 
 ![License](https://img.shields.io/badge/license-Apache%202.0-blue)
 
 ## Table of Contents
 
 - [Introduction](#introduction)
 - [Features](#features)
-- [Getting Started](#getting-started)
-  - [Prerequisites](#prerequisites)
-  - [Installation](#installation)
+- [Prerequisites](#prerequisites)
+- [Installation](#installation)
 - [Usage](#usage)
-- [Command-line Arguments](#command-line-arguments)
+- [Command-line Options](#command-line-options)
+- [Run Examples](#run-examples)
 - [Output](#output)
+- [Contributing](#contributing)
 - [License](#license)
+- [Acknowledgments](#acknowledgments)
 
 ## Introduction
 
-The Video Object Detection and Tracking Script is a Python tool designed for processing videos or sequences of images to detect and track changes between frames. It employs various image processing techniques and algorithms to identify differences, track objects, and manage stationary objects. This script is useful in scenarios where you need to analyze and visualize motion patterns or stationary periods within a video stream.
+The  is a Python tool designed for processing videos or sequences of images to detect and track changes between frames. It employs various image processing techniques and algorithms to identify differences, track objects, and manage stationary objects. This script is useful in scenarios where you need to analyze and visualize motion patterns or stationary periods within a video stream.
 
 ## Features
 
@@ -26,16 +28,14 @@ The Video Object Detection and Tracking Script is a Python tool designed for pro
 - Provides command-line interface for customization and control.
 - Generates output videos and log files to visualize and analyze results.
 
-## Getting Started
-
-### Prerequisites
+## Prerequisites
 
 - Python 3.6 or higher
 - OpenCV (cv2)
 - NumPy (np)
 - scikit-image (skimage)
 
-### Installation
+## Installation
 
 1. Clone the repository:
 
@@ -55,38 +55,92 @@ To use the Video Object Detection and Tracking Script, follow these steps:
 
 1. Ensure that you have Python and the required libraries installed (see Prerequisites).
 
-2. Place your input video or image sequence in the `./data/videos/` directory.
-
-3. Run the script using the following command:
+2. Arrange your input folder and output folder accordingly. Recommended folder arrangement:
+
+```
+Object-Detection-Camera-Feed/
+   main.py
+   preprocess.py
+   processor.py
+   tracker.py
+   ignore.txt
+   data/
+      videos/
+         vid.mp4
+         ...
+      masks/
+         mask.jpg
+         ...
+   output/
+      ...
+```
+
+3. Modify the `ingore.txt` file as you wish. 
+
+- This text file will ignore the box location that you want. 
+- Keep in mind that it will only ignore objects that have matching or bigger `iou` with the boxes inside `ingore.txt`.
+- Format is one `[x1, y1, x2, y2]` per line.
+
+4. Change directory to the project folder then run the program using the following command:
 
    ```bash
-   python video_object_detection_tracking.py -i ./data/videos/your_video.mp4
+   python main.py -i your_video.mp4
    ```
 
-4. Customize the behavior using the available command-line arguments (see [Command-line Arguments](#command-line-arguments)).
+5. Customize the program's behavior using the available command-line arguments (see [Command-line Arguments](#command-line-arguments)).
 
-5. The processed video and log files will be saved in the `./output/` directory.
+6. Customize preprocess variables accordingly to your input video by setting `--white` and `--black` threshold to optimize the result.
 
-## Command-line Arguments
+7. The processed video and log files will be saved in the `./output/` directory or the path you provided
+
+## Command-line Options
 
 - `-i`, `--input`: Path to the input video or image sequence. (required)
 - `-o`, `--output`: Path to the directory where outputs will be saved.
 - `-m`, `--mask`: Path to a mask image to exclude specific areas from detection.
-- `-u`, `--iou`: Intersection over Union (IoU) threshold for matching object positions.
-- `-r`, `--refresh`: Number of frames between stationary object checks.
-- `-g`, `--ignore`: Path to the file containing ignored box locations (one per line).
-- `-t`, `--track`: Enable tracking of detected objects (True or False).
-- `-b`, `--bsub`: Enable background subtraction (True or False).
-- `-k`, `--outmask`: Save masked video (True or False).
+- `-g`, `--ignore`: Path to a list of positions of boxes that will be ignored.
+- `--iou`: Intersection over Union (IoU) threshold for matching object positions.
+- `--min-size`: Minimun area of the contour box to be recorded as an object.
+- `--track-rate`: Number of frames between stationary object checks.
+- `--white`: Set minimum value (from 0 to 255) to be white pixel otherwise will be turned black.
+- `--black`: Set the minimum value (from 0 to 255) to be black pixel otherwise will be turned white.
+- `--gray`, `--no-gray`: Turn on or off grayscale convertion in preprocessing.
+- `--contrast`, `--no-contrast`: Turn on or off auto contrast in preprocessing.
+- `--blur`, `--no-blur`: Turn on or off blurring in preprocessing.
+- `--edge`, `--no-edge`: Turn on or off edge detection in preprocessing.
+- `--save`, `--no-save`: Turn on or off saving result video.
+
+## Run Examples
+
+1. Basic:
+  
+```bash
+python main.py -i './input_images' -o './output_images'
+```
+
+2. Run with modified stationary object tracking rate and higher iou:
+
+```bash
+python main.py -i './vid.mp4' -o './output' --track-rate 15 --iou 0.9
+```
 
 ## Output
 
-The script generates the following output:
+The script generates the following output to the path of your choice:
 
-- `./output/output.mp4`: Processed video with detected and tracked objects.
+- `./output/result_video.mp4`: Processed video with detected and tracked objects.
 - `./output/log.txt`: Log file containing information about detected stationary objects.
-- `./output/snapshot.jpg`: Snapshot of all the objects detected as new stationary object
+- `./output/id_[id_number].jpg`: Snapshot of all the objects detected as new stationary object
+
+## Contributing
+
+Contributions are welcome! If you have any suggestions or improvements for this code, please feel free to submit a pull request.
 
 ## License
 
-This project is licensed under the [Apache License 2.0](LICENSE).
+This project is licensed under the [Apache License 2.0](LICENSE).
+
+## Acknowledgments
+
+- OpenCV: https://opencv.org/
+- The scikit-image library: https://scikit-image.org/

ignore.txt

@@ -0,0 +1,2 @@
+[1544, 506, 1661, 694]
+[1539, 505, 1648, 695]

main.py

@@ -27,12 +27,12 @@ def main():
     parser.add_argument('-i', '--input', type=str, help='Path to the input video.', required=True)
     parser.add_argument('-o', '--output', type=str, help='Path to output folder.', default=VideoProcessor.DEFAULT_OUT_PATH)
     parser.add_argument('-m', '--mask', type=str, help='Path to a mask image.', default=None)
-    parser.add_argument('--ignore', type=str, help='Path to a list of positions of boxes in the frame that you want the program to ignore. Each line contains 1 box position. Example format: [x1, y1, x2, y2]', default=None)
+    parser.add_argument('-g', '--ignore', type=str, help='Path to a list of positions of boxes in the frame that you want the program to ignore. Each line contains 1 box position. Example format: [x1, y1, x2, y2]', default=None)
     parser.add_argument('--iou', type=float, help='IOU threshold for object matching.', default=Tracking.DEFAULT_IOU_THRESHOLD)
     parser.add_argument('--min-size', type=int, help='Minimun area of the contour box to be recorded as an object.', default=Tracking.DEFAULT_MIN_SIZE)
-    parser.add_argument('--track-rate', type=int, help='Tracking rate for stationary object detection.', default=Tracking.DEFAULT_TRACK_RATE)
-    parser.add_argument('--white', type=int, help='Determine the minimum value to be white pixel otherwise will be turned black.', default=VideoProcessor.DEFAULT_WHITE_THRESHOLD)
-    parser.add_argument('--black', type=int, help='Determine the minimum value to be black pixel otherwise will be turned white.', default=VideoProcessor.DEFAULT_BLACK_THRESHOLD)
+    parser.add_argument('--track-rate', type=int, help='Number of frames between stationary object checks', default=Tracking.DEFAULT_TRACK_RATE)
+    parser.add_argument('--white', type=int, help='Set the minimum value (from 0 to 255) to be white pixel otherwise will be turned black.', default=VideoProcessor.DEFAULT_WHITE_THRESHOLD)
+    parser.add_argument('--black', type=int, help='Set the minimum value (from 0 to 255) to be black pixel otherwise will be turned white.', default=VideoProcessor.DEFAULT_BLACK_THRESHOLD)
     parser.add_argument("--gray", type=bool, help="(bool) False to turn off grayscale in preprocessing, True otherwise", action=argparse.BooleanOptionalAction, default=True)
     parser.add_argument("--contrast", type=bool, help="(bool) True to turn on auto contrast in preprocessing, False otherwise", action=argparse.BooleanOptionalAction, default=False)
     parser.add_argument("--blur", type=bool, help="(bool) True to turn on blurring in preprocessing, False otherwise", action=argparse.BooleanOptionalAction, default=False)

processor.py

@@ -200,5 +200,4 @@ def process_video(self):
 
         self.video.release()
         # result_video.release()
-        cv2.destroyAllWindows()
-        
+        cv2.destroyAllWindows()