Created a vehicle detection and tracking pipeline with OpenCV, histogram of oriented gradients (HOG), and support vector machines (SVM). Optimized and evaluated the model on video data from a automotive camera taken during highway driving.

MEDIUM: VEHICLE TRACKING

Version 2 uses the following parameters for feature extraction

feature_params = {
  'color_model': 'yuv',                # hls, hsv, yuv, ycrcb
  'bounding_box_size': 64,             # 64 pixels x 64 pixel image
  'number_of_orientations': 11,        # 6 - 12
  'pixels_per_cell': 16,               # 8, 16
  'cells_per_block': 2,                # 1, 2
  'do_transform_sqrt': True
}
# [3 x 3 block positions] x [2 x 2 cells per block] x [11 orientations] x [3 channels] = 1,188 features

Version 1 uses the following parameters for feature extraction

feature_params = {
  'color_model': 'hls',                # hls, hsv, yuv, ycrcb
  'bounding_box_size': 64,             # 64 pixels x 64 pixel image
  'number_of_orientations': 12,        # 6 - 12
  'pixels_per_cell': 8,                # 8, 16
  'cells_per_block': 2,                # 1, 2
  'do_transform_sqrt': True
}

# [7 x 7 block positions] x [2 x 2 cells per block] x [12 orientations] x [3 channels] = 7,056 features

SAMPLE USAGE

To run any notebook properly, copy the jupyter notebooks from the /ipynb folder to the root directory. This is so that each notebook sees relevant files, the most relevant files being the python classes.

• classifier_training.ipynb
• feature_sourcer_test.ipynb
• classifier_test.ipynb
• slider_test.ipynb
• heatmap_test.ipynb
• pipeline.ipynb

CLASSES

• featuresourcer.py
• binaryclassifier.py
• slider.py
• heatmap.py

VERSION 1

Data from From Kitti Benchmark

• Non-Vehicle Data
• Vehicle Data