run.py: borrow from https://github.com/csbanon/canny-edge-detectorcanny.py: working with OpenCVHeatmap_Canny.py: Both Canny Edges Estimation & Heatmap Extraction
Canny Edge Estimation is an implementation of the edge detection method defined by John Canny in 1986. This is done by performing the following computational steps:
- Perform Gaussian smoothing on the image.
- Compute the gradient magnitude and orientation using derivative masks.
- Perform non-maximum suppression.
- Perform hysteresis thresholding.
The input of this program is the file path to an image of your choice.
Heatmap Estimation is simply a lightweight Fully Convolutional Network (but the architectural is more like SegNet), including:
- 3 Conv layers
- 3 deConv layers