ChipAnalyser is an image-processing program designed to analyze machining chips from a series of images or a video file. It measures the following characteristics of the chip in each image:
- Tool-chip contact length
- Average peak thickness
- Average valley thickness
Measurement results are saved in a CSV file. Additionally, the program can generate graphical renderings to illustrate the measurement process:
python ChipAnalyser [-S] -i INPUT_IMAGES -o OUTPUT_DIRECTORY [-s SCALE] [-b BATCH_SIZE] [-r]-S, --silent: if this option is enabled, the program does not display progress bar
-i INPUT_IMAGES: path to the directory containing the images to be analyzed, or path to the video whose images are to be analyzed.
-o OUTPUT_DIRECTORY: path to the directory where output files will be written
-s SCALE: length of a pixel in µm (1 by default).
-b BATCH_SIZE: size of the input image batches (10 by default).
-e: if this option is enabled, the program produces graphical renderings to explain how the feature are extracted, else, no rendering is done and the program simply extracts the features from the inputs.
- NumPy: Perform vectorized operation on large arrays.
- OpenCV: Implementation of famous image processing algorithms.
- SciPy: Signal processing to find peaks and valleys after chip thickness measurement.
- Scikit-image: Read input image files. Provides a line rasterization algorithm used to detect the inner contour of the chip by ray casting.
- Scikit-video: Read input video file. Write output video files.
- Matplotlib: Generate output plots.
# create and activate the virtual environment
conda create -n machining-chip-analysis python=3.11
conda activate machining-chip-analysis
# install dependencies inside the virtual environment
conda install "numpy<1.24" opencv -c conda-forge
conda install matplotlib scikit-video -c conda-forge
conda install scipy scikit-image -c anacondapip install "numpy<1.24" opencv-python scipy matplotlib scikit-image scikit-video