Convolution-Based Edge Detection for Image/Video in Block DCT Domain
✍ Scribed by B. Shen; I.K. Sethi
- Publisher
- Elsevier Science
- Year
- 1996
- Tongue
- English
- Weight
- 752 KB
- Volume
- 7
- Category
- Article
- ISSN
- 1047-3203
No coin nor oath required. For personal study only.
✦ Synopsis
tent descriptors for images and videos. Previous work in compressed domain edge feature extraction only provided This paper presents a scheme for performing convolution operation directly on compressed images without decomcertain coarse interpretation of statistical features, which pressing them first. The use of such a scheme is demonstrated tend to ignore important low-level features such as edges, and discussed by showing the implementation of the Laplaciancorners, or fine textures. There have been some prelimiof-Gaussian operator for edge detection. We present a complete nary edge extraction methods based on the classification evaluation of the different parameters involved in this process of DCT coefficients. Some coarse edges can be extracted and show edge detection results on several real images through out directly from transform coefficients; e.g., Arman et al.
our proposed scheme. In each case, it is shown that the proposed
[2] used the number of nonzero coefficients in DCT block scheme of directly performing convolution on the compressed to decide whether an edge exists. Shen and Sethi [17] used data leads to not only a significant computation speedup but patterns of DCT coefficients to decide edge parameters also yields better edges.