Development of edge detection algorithm for blocky compressed image
Nowadays, compressed image is a very useful way to transfer or store data because of its smaller size in comparison to the original image. However, if the compression ratio is high, the produced image may be corrupted by blocky artifacts. In many applications that require only the detection of the i...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/4520/ http://eprints.uthm.edu.my/4520/1/DEVELOPMENT_OF_EDGE_DETECTION_ALGORITHM.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Nowadays, compressed image is a very useful way to transfer or store data because of
its smaller size in comparison to the original image. However, if the compression
ratio is high, the produced image may be corrupted by blocky artifacts. In many
applications that require only the detection of the image edges, it is difficult to
distinguish the edges, fine details and noise. This task is more difficult in compressed
image, since the generated blocky artifacts have almost the same properties as the true
edges. Therefore, in this study we develop an algorithm that could detect only the
edges from a compressed image. For horizontal direction, two variables are calculated
for every pixel in the odd-numbered rows of the image. The two variables are the
discontinuities in horizontal pixel intensity gradient at the neighbouring pixels, which
will be compared with a threshold value. If both values are higher than the threshold
value, the corresponding pixel is classified as edge. Detection in vertical direction is
performed in the same manner. The detection results for horizontal and vertical
directions are generated in two new black and white images, respectively. The
generated new images are then combined to produce the final detected image. From
our simulation, we found that this detection method has successfidly detected edges
without including the blocky artifacts and fine details. |
|---|