Object Boundary Detection using Neural Network in Deep Learning
S.P.Maniraj1, Sreenidhi G2, Peddamallu Sravani3, Yeshwant Reddy4
1S.P.Maniraj, Assistant Professor , Srm Institute Of Science And Technology , Chennai, India.
2Sreenidhi G, UG Scholar, Srm , Institute Of Science And Technology , Chennai, India.
3Peddamallu Sravani, UG Scholar, Srm Institute Of Science And Technology , Chennai, India.
4Yeshwant Reddy, UG Scholar , Niit University, Rajasthan, India.
Manuscript received on September 23, 2019. | Revised Manuscript received on October 15, 2019. | Manuscript published on October 30, 2019. | PP: 4453-4457 | Volume-9 Issue-1, October 2019 | Retrieval Number: A1608109119/2019©BEIESP | DOI: 10.35940/ijeat.A1608.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Object boundary detection is a new trend where we apply them in all aspects like automated self-cars, domestic robots and much more. . In traditional methods, the two objectives of object boundary detection for the given image patch as follows, mainly the first part will generate the partially segmented output, where the other part of the system will generate output as a block outline being on each of the object in the input environment. . It will generate a set of object proposals and apply CNN (Convolutional Neural Network) in the proposed area to limit the computing time and get accurate results. After preprocessing stage, it will reach the VGG(Visual Geometry Group)without fully connected layer. Duration test time, our model depends on entire test image and it will generate a set of segmentation. Detection of objects in an image and making outline around the edge of the boundary of object by means of machine learning algorithms. Hence project is not likely to be like other proposals related to low level edge segmentation and our model will improve its clarity while training it or put in use continuously. Thus the improvements can be felt when its put to use it show more clarity than before depending on how many time you use.
Keywords: Deep Learning, CNN, VGG, Deep mask, segmentation mask.