Endomicroscopy Image Recognition using Ensemble Neural network with Contrast Limited Adaptive Histogram Equalisation
Kalaivani. N1, Kani Mozhi. N2, Kanimozhi. M3, Kalieswari. S4, Kuralarasi. R5
1Kalaivani.N*, Electronics and Communication Engineering, Sri Krishna College of Engineering and Technology, Coimbatore, India.
2KaniMozhi.N, Electronics and Communication Engineering, Sri Krishna College of Engineering and Technology, Coimbatore, India.
3KaniMozhi.M, Electronics and Communication Engineering, Sri Krishna College of Engineering and Technology, Coimbatore, India.
Manuscript received on April 05, 2020. | Revised Manuscript received on April 25, 2020. | Manuscript published on April 30, 2020. | PP: 44-48 | Volume-9 Issue-4, April 2020. | Retrieval Number: C6438029320/2020©BEIESP | DOI: 10.35940/ijeat.C6438.049420
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: Endomicroscopy is a small tool used for cancer diagnosis, this enables in-vivo imaging at microscopic resolution closely to histology image during endoscopic procedures and captured image within the dataset has high imaging quality resulting in an inequality between moral and poor-quality images. There’s no clear demonstration of the artifacts in an endomicroscopy producer. During this proposed method, the ensemble neural network (ENN) approach models to scale back the variance of predictions and reduce generalization error with contrast limited adaptive histogram equalization (CLAHE) algorithm were used to recover the image pixel balancing. Binary classification of accuracy 98.79% has been achieved.
Keywords: Breast cancer, Ensemble neural network (ENN), Contrast limited adaptive histogram equalization (CLAHE).