Defect Detection On Manufacturing Product Images by Applying Weekly Detector Algorithm using Deep Learning Technology

  • Dr. S.Praveen Kumar , Dr.K.Naveen Kumar, Dr.Y.Srinivas, Dr.G.V.S Rajkumar

Abstract

Deep learning permits process models that square measure created out of diverse handling layers to find
out illustration of data with various levels of abstraction. The paper is especially aimed towards making
Associate intending innovative pc vision program that will act as Associate in aid to work out the
defective surfaces on any product. To realize such machine-controlled approaches, it's needed to utilize
one in every of the many Deep Learning Algorithms that exist now a days. VGG net was chosen for
assessing the desired results.
Additionally, the scale of information set was augmented to confirm that the trained model doesn't over
match the data. As a result, four totally different classifications were obtained once the complete dataset
is fed into the trained model.For segmentation we have a tendency to enforced feeble detector to induce
areas with defects to be thermally known and localize their class-specific areas from the category
activation maps. This paper demonstrates that our network is in a position to classify and localize the
discriminative areas on the defects of the photographs fed.

Published
2020-05-02
How to Cite
Dr. S.Praveen Kumar , Dr.K.Naveen Kumar, Dr.Y.Srinivas, Dr.G.V.S Rajkumar. (2020). Defect Detection On Manufacturing Product Images by Applying Weekly Detector Algorithm using Deep Learning Technology. International Journal of Advanced Science and Technology, 29(7), 186 - 194. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/13206
Section
Articles