i. Laplacian is a second derivative mask, which can be used for edge detection.

ii. However, in its original form as lapalacian is a second derivative mask, it is very sensitive to noise. Thus if an image contains noise, the laplacian gives very large values and also ruins the image in the process.

iii. The magnitude of laplacian produces double edges, which is an undesirable property.

iv. Thus laplacian alone is never used for edge detection

v. But, if some kind of noise filtering is done prior to the laplacian operation, better results can be obtained..

vi. Thus we conclude that Laplacian, on its own, is not a good edge detector, however with noise being filtered out prior to using Laplacian, better edge detection can be obtained.