PodCounter

PodCounter is an automated image processing tool designed to count pods in captured photos, correct distortions, crop backgrounds, rename images based on QR codes, and generate a report with the total number and average length of pods. It is tailored for accurate and consistent analysis of agricultural research images.

Requirements

To run this project, you need the following Python libraries:

• opencv-python
• numpy
• tqdm
• matplotlib

You can install all dependencies using the requirements.txt file:

pip install -r requirements.txt

Features

1. Fisheye Distortion Correction
   Corrects distortion in images captured with GoPro Hero 4 cameras.

2. Background Cropping
   Crops the background outside the board area in images. The code is adjusted to work with blue boards.

3. Image Renaming
   Renames images based on detected QR codes.

4. Pod Counting
   Detects and counts pods in the images, calculates their lengths using a white label as a size reference (can be the QR code label), and saves each individual length in the pods.csv file. In the processed images, the pods are outlined in red, while the white labels are outlined in green to facilitate visual verification. Each pod is assigned a unique identifier (ID) at its centroid, enabling the individual identification of each pod in the pods.csv file.

5. Summary Generation
   Creates a summary table with the total number of pods and the average pod length for each processed image, and save in the summary_table.txt file.

How to Use

Taking Pictures

To ensure accurate pod counting and measurement, please follow these guidelines when capturing images:

1. Use a Matte Blue Board

• Use a flat, matte blue board as the background.
• This color provides optimal contrast for pod detection.

2. Include a White Reference Object

• Place a white object with a known area beneath the blue board. This polygon should:
  • Be the largest white object in the image.
  • Optionally can be a label contain a QR code to help identify the sample.

3. Position the Pods Properly

• Spread the pods with sufficient space between them.
• Important: Pods that touch or are too close to each other will be detected and counted as a single pod.
• Avoid stacking or overlapping the pods.

4. Keep the Board Clean

• Do not place any objects on the blue board other than the pods and the white reference polygon.
• Remove any debris, shadows, or other obstructions that could interfere with image processing.

5. Reference Example Image

• For guidance on proper setup, refer to the example image located at:
  example/pods_counting/counted/314_MB-DOP_1.jpg.
• This example shows:
  • How the pods and the white reference object should be positioned.
  • What happens when pods touch (e.g., pods 4, 22, and 55 in the example are incorrectly counted as a single pod).

By following these guidelines, you ensure that the images are optimized for processing by PodCounter.

Setup

1. Organize the images into an input folder.
2. Set the arguments for the input and output directories in main.py:
   • folder_path: Path to the folder containing the input images.
   • counting_path: Path to save the processed images and summary.
   • label_area_cm2: Defines the area of the white label in square centimeters.
3. Run the main script main.py:

python main.py

Output

• Background Cropped Images: Saved in the counting_path directory.
• Images with Pods Highlighted: Saved in the counting_path/counted directory.
• Individual Pods Size: Saved in the counting_path/counted/pods.csv.
• Summary: Generated at counting_path/counted/summary_table.txt.

Step-by-Step Guide for Beginners

If you're new to Python, follow these instructions to set up and use PodCounter:

1. Install Python

1. Download Python from the official website: python.org.
2. During installation, make sure to:
   • Check the box that says "Add Python to PATH" before clicking "Install Now."
   • Complete the installation by following the prompts.

2. Install a Code Editor (Optional)

• For a better experience, download and install a code editor like Visual Studio Code.

3. Download PodCounter

1. Download the PodCounter project from the GitHub repository:
   • Click "Code" on the repository page and select "Download ZIP". Extract the ZIP file to a folder of your choice.

4. Install Required Python Libraries

1. Open a terminal or command prompt.
   • On Windows: Press Win + R, type cmd, and hit Enter.
   • On macOS/Linux: Open the Terminal application.
2. Navigate to the project folder where the requirements.txt file is located:

   cd path/to/PodCounter

3. Install the required libraries:

   pip install -r requirements.txt
   

5. Prepare Your Images

1. Organize the images you want to process in a folder (e.g., input_images).
2. Ensure that the images follow the guidelines in the Taking Pictures section.

6. Configure the Script

1. Open the main.py file in a text editor or code editor.
2. Set the following variables:
   • folder_path: The path to your input folder containing the images.
   • counting_path: The path where processed images and reports will be saved.
   • label_area_cm2: The area (in square centimeters) of the white reference polygon beneath the blue board.

7. Run PodCounter

1. In the terminal or command prompt, navigate to the project folder:

   cd path/to/PodCounter

2. Run the script:

   python main.py
   

8. Check the Results

1. Processed images and reports will be saved in the counting_path directory:
   • Background Cropped Images: Found directly in the directory.
   • Images with Pods Highlighted: Saved in the counting_path/counted folder.
   • Pod Sizes: Saved in the pods.csv file.
   • Summary Report: Saved in the summary_table.txt file.