Video-Based Detection of Squint and Cataract for Accessibility-Aware Adaptive Web Interface Rendering

Jul 16, 2026

Upload by Eden

Squint and cataract significantly impair visual perception and user interaction capability. The paper presents a real-time, video-based automated detection system for both conditions using computer vision and image processing. A media-pipe face-mesh model with 478 facial landmarks extracts geometric ocular features to support multi-class squint classification, while cataract presence and severity are estimated through grayscale intensity and histogram-based lens opacity analysis. Short video sequences recorded with standard laptop or mobile cameras enable low-cost, large-scale deployment, achieving 98.39% accuracy for squint detection and 96.90% cataract classification. The framework is designed to support accessibility-aware visual impairment inference for future adaptive web and interface systems.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

Introduction

Literature Review

VIDEO-BASED DETECTION OF SQUINT AND CATARACT FOR ACCESSIBILITY-AWARE ADAPTIVE WEB INTERFACE RENDERING Amar Ranjan Dash 1 and Manas Ranjan Patra 2 1Department of Computer Science, Berhampur University, Berhampur, India 2CSE Department, NIST University, Pallur Hills, Berhampur, India ABSTRACT Squint and cataract are major ocular disorders that majorly affect visual perception and interaction capability. This paper proposes a real-time video-based automated detection system for squint and cataract detection based on computer vision and image processing methods. The proposed system uses a media-pipe face-mesh (a 478-point facial landmark detection model) to extract geometric ocular features for multi-class squint classification. Simultaneously, The presence and severity cataract is estimated through grayscale intensity and histogram-based lens opacity analysis. The system records short video sequences with standard laptop or mobile cameras, which can be deployed at low costs and on a large scale. The experimental performance has shown great accuracy in the detection of squint (98.39%) and classification of cataract (96.90%). Besides automatic ocular analysis, the proposed framework is also made accessible for visual impairment inference which will be integrated with future adaptive userinterface and Web accessibility systems for people with visual impairment. KEYWORDS Squint Detection, cataract Detection, Video Mining, Facial landmark, Pupil tracking, Digital image processing, Computer vision. 1. INTRODUCTION The eye is a sense organ which processes light and perceives visual and depth information. Any structural or functional defect of the eye can cause visual deficiency like squint, cataract, blur and colour blindness. Squint occurs due to abnormal ocular alignment. It means one or both eyes deviate from their anatomically coordinated positions. Squint results in impaired binocular vision and reduced depth perception. Cataract develops when crystallin proteins begin to clump together in the lens causing the lens to become opaque, causing problems with vision, glare and contrast perception. Typical diagnosis of the disorders involves specialized ophthalmic equipment and expert clinical evaluation, which are not readily available in low-resource areas. The application of automated techniques such as image and video analysis are now possible for the identification of ocular disease, due to recent advances in computer vision and medical image processing. However, most existing approaches focus on isolated disease detection using static images and provide limited support for real-time multi-condition analysis. This paper presents a hybrid approach for the simultaneous detection of squint and cataract based on geometric and intensity based analysis of the various features of the eye from live video. Media-pipe face-mesh (a 478-point facial landmark model) is used to locate the pupils and classify the image into multiple classes of squints using the facial landmark model, and the severity of cataract in the treated image is estimated by grayscale histogram and lens opacity analysis. The proposed system operates in real time using standard cameras. Additionally it supports accessibility-aware visual impairment inference for adaptive human-computer interaction systems. 2. LITERATURE REVIEW Squint occurs when two pupils are misaligned and look in different direction, preventing coordinated binocular vision. Squint occurs when one pupil deviates inward, outward, upward or downward relative to the other resulting in misalignment and impaired binocular vision. First A RElkington and P T Khaw [01] have analyzed different type of squint and mentioned different type of clinical test need to be conducted for detection of six type of squint. S. Yadav [02] have developed an algorithm for analysing age related macular degeneration in retina. G. Lennerstrand [03] have analyze the structure and function of eye muscles. He ha

FAQs

How does the system detect squint from a video?

It uses a media-pipe face-mesh model to extract 478-point facial landmarks and derive geometric ocular features, then classifies the eye alignment into multiple squint categories in real time.

How is cataract severity estimated in the proposed approach?

Cataract presence and severity are inferred from grayscale intensity and histogram-based lens opacity analysis of the eye region.

What accessibility-related capability does the framework support?

The system produces visual impairment inference that can be integrated into future adaptive user-interface and web accessibility systems for people with visual impairment.

Jul 16, 2026

Upload by Eden