filter2D

Convolves an image with the kernel.

cv2.filter2D(src, ddepth, kernel[, dst[, anchor[, delta[, borderType]]]]) → dst

• src (np.ndarray): Input image.
• ddepth (cv2.CV_*): Output image depth. Default is -1 to use src1.depth(), and this is used in this app.
• kernel (np.ndarray): Convolution kernel (or rather a correlation kernel), a single-channel floating point matrix; if you want to apply different kernels to different channels, split the image into separate color planes using split() and process them individually.
• dst (optional; np.ndarray): Output image of the same size and type as src.
• anchor (optional; tuple: (int, int)): Anchor of the kernel that indicates the relative position of a filtered point within the kernel. The tuple is provided as (column, row), and the anchor should lie within the kernel. Default value (-1,-1) means that the anchor is at the kernel center.
• delta (optional; float): Optional value added to the filtered pixels before storing them in dst.
• borderType (optional; cv2.BORDER_*): Pixel extrapolation method. Default is BORDER_DEFAULT. Choose from:
  • BORDER_CONSTANT
  • BORDER_REPLICATE
  • BORDER_REFLECT
  • BORDER_REFLECT_101 or BORDER_REFLECT101 or BORDER_DEFAULT
  • BORDER_ISOLATED

The function applies an arbitrary linear filter to an image. In-place operation is supported. When the aperture is partially outside the image, the function interpolates outlier pixel values according to the specified border mode. The function uses the DFT-based algorithm in case of sufficiently large kernels (around \(11 \times 11\) or larger) and the direct algorithm (that uses the engine retrieved by createLinearFilter()) for small kernels.

• The function does actually compute correlation, not the convolution: $$\text{dst}(x,y) = \sum_{\substack{0 \leq x' < \text{kernel.cols} \text{ and } 0 \leq y' < \text{kernel.rows}}} \text{kernel}(x', y') * \text{src}(x+x' - \text{anchor}.x, y+y'-\text{anchor}.y)$$ That is, the kernel is not mirrored around the anchor point.
• If you need a real convolution, flip the kernel using flip and set the new anchor to (kernel.cols - anchor.x - 1, kernel.rows - anchor.y - 1).
• BORDER_WRAP and BORDER_TRANSPARENT are not supported.
• If you want to use BORDER_CONSTANT, you should manually add border to your image with copyMakeBorder, otherwise the default value is set to 0.

• OpenCV Documentation