Sie sind auf Seite 1von 3

ASSIGNMENT 1

Results & Observation

1. Apply Gaussian blur on the image to reduce noise.


Result: Gaussian Filter is applied on input image with different sigma value (sigma=0,
sigma=1, sigma=3) to reduce the noise.

Input image output image

Observation: Increasing the standard deviation continues to reduce/blur the intensity of the
noise, but also attenuates high frequency detail (e.g. edges) significantly.
2. Implement Canny Edge detector.
Result: Canny Edge Detector is implemented with different minVal and maxVal as
Hysteresis Thresholding.

Input image output image

Observation: keeping minVal same and increasing maxVal 100 to 300 as shown above in
image increases the sharp edge detection in image. So setting the correct range of minVal and
maxVal help in getting correct edge detection in the input image.

3. Implement Harris corner detector.


Result: Harris Corner detector applied on input image and corner detected in output image
marked as square.
Input image output image

Input image output image

Observation: Since corners represents a variation in the gradient in the image. We are going
to sweep a window W(x,y) (with displacements in the x direction and in the right
direction) and will calculate the variation of intensity. A score is calculated for each window,
to determine if it can possibly contain a corner a window with a score greater than a certain
value is considered a “corner” and marked as square in output image. Above result is obtained
by setting u=2,v=3 and score =0.04.

Das könnte Ihnen auch gefallen