Development of Scale Invariant lens Opacity Estimation System using Hough Circle Detection Transform, Normalization and Entropy
Amol Jagadale1, Santosh Sonavane2, Dattatray Jadhav3
1A. B. Jagadale*, Department of Electronics, RSCOE, Tathawade, Savitribai Phule Pune University, Pune, working as Asst Prof. at SKNSCOE, Korti, Pandharpur, Maharashtra, India.
2S. S. Sonavane, Director, Symbiosis Skills & Professional University (SSPU) Pune, Maharashtra, India.
3D. V. Jadhav, DTE, Pune, Maharashtra, India.
Manuscript received on January 03, 2021. | Revised Manuscript received on March 11, 2021. | Manuscript published on March 30, 2021. | PP: 8-10 | Volume-10 Issue-5, March 2021 | Retrieval Number: 100.1/ijitee.E86140310521| DOI: 10.35940/ijitee.E8614.0210421
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: Clear eye lens is responsible for correct vision. Ageing effect acquires opacity at lens structure causing foggy or blurred vision. It is termed as cataract. This may become cause of permanent blindness if remain unidentified and untreated. Due to hazards change in environment and adoption of sluggish lifestyle many diseases like cataract are becoming universal challenge for health organization over the world. Lack of medication and diagnosis facility in developing countries makes cataract as savior vision problem. Proposed methodology suggests image processing based, low cost solution for lens opacity or cataract detection. In this system eye lens image from input image is acquired using Iterative Hough circle detection transform. It is normalized using Daugman’s rubber sheet normalization algorithm which makes system scale invariant. Structural variation in normalized lens image is estimated in terms of entropy or mean value. Comparison of right and left half entropies of normalized image is basis for estimation of lens opacity. It is used to detect and categorize lens opacity or cataract. This system easily categorize lens opacity based on structural features of opacity in one of three grades such as “No cataract”, “Cortical cataract” or “Nuclear cataract”.
Keywords: Cataract, Hough circle detection transform, Daugman normalization, Entropy, Structural features.