Advanced Sketch Based Image Retrieval System Using Object Boundary Selection Algorithm
Aishwarya A. Deore, Baisa Laxman Gunjal · Zenodo (CERN European Organization for Nuclear Research) · 2016
In recent era for some cases, colorful image retrieval on the basis of their sketches is area of interest as far as image processing is concern. In image retrieval process sketch of particular image is given as an input and its colorful image is expected as an output. Problem of sketch and actual image matching is that there is appearance gap between them. This gap is due to noise in edges of photo realistic images. Hence our proposed process considers these issues in Sketch Based Image Retrieval process. Our system eliminates the impact of noise in edges of photo realistic images. Our proposed process is based on histogram of line relationship (HLR) descriptor. We proposed "Object boundary selection" algorithm that followed by HLR. In our system we mainly contributing that, matching process is done on Hadoop using Map-reduce technique.