C4W1L03 More Edge Detection

What distinguishes a "positive edge" from a "negative edge" in image processing?

• Positive edges represent light-to-dark transitions, while negative edges represent dark-to-light transitions.

• Positive edges represent dark-to-light transitions, while negative edges represent light-to-dark transitions.

• Positive edges are horizontal, while negative edges are vertical.

• Positive edges are strong, while negative edges are weak.

When an image with a vertical edge transitioning from light to dark on the left is convolved with a vertical edge detection filter, what kind of values are typically produced in the output matrix at the edge location?

• Negative values

• Zero values

• Positive values

• Random values

If an image has a vertical edge transitioning from dark to light on the left, and it's convolved with the same vertical edge detection filter used for light-to-dark transitions, what will be the typical values in the output matrix at the edge?

• Positive values

• Negative values

• Zero values

• Unchanged values

To detect vertical edges regardless of whether they are light-to-dark or dark-to-light transitions, what operation can be applied to the output of the convolution?

• Squaring the values

• Taking the absolute value

• Multiplying by -1

• Adding a constant bias

A vertical edge detection filter, such as the one with columns [1, 1, 1], [0, 0, 0], [-1, -1, -1], is designed to detect what kind of pixel intensity change?

• Bright pixels on top and dark pixels on the bottom.

• Dark pixels on top and bright pixels on the bottom.

• Bright pixels on the left and dark pixels on the right.

• Dark pixels on the left and bright pixels on the right.

A horizontal edge detection filter, such as the one with rows [1, 1, 1], [0, 0, 0], [-1, -1, -1], is designed to detect what kind of pixel intensity change?

• Bright pixels on the left and dark pixels on the right.

• Dark pixels on the left and bright pixels on the right.

• Bright pixels on top and dark pixels on the bottom.

• Dark pixels on top and bright pixels on the bottom.

Besides the basic vertical edge detection filter (1 0 -1, 1 0 -1, 1 0 -1), what are two other commonly used hand-coded filters mentioned for edge detection?

• Gaussian and Laplacian filters

• Median and Bilateral filters

• Sobel and Scharr filters

• Prewitt and Canny filters

In the context of deep learning, how are the parameters (weights) of an edge detection filter typically determined for optimal performance?

• They are hand-coded by computer vision researchers.

• They are randomly initialized and remain fixed.

• They are learned automatically from data using backpropagation.

• They are pre-defined by the image format.