Development and Validation of a Digital Image Processing-Based Pill Detection Tool for an Oral Medication Self-Monitoring System

• DOI: 10.20944/preprints202202.0204.v1
• Publication Date: 2022-02-17T07:56:36Z
• Authors: Jannis Holtkötter; Rita Amaral; Rute Almeida; Cristina Jácome; Ricardo Cardoso; Ana Pereira; Mariana Pereira; Ki Chon; João Almeida Fonseca
• Keywords: Blisters; Pill; Hough Transform; Digital Image Analysis

Objective tools to track medication adherence are lacking. A tool monitor pill intake that can be implemented in mHealth apps without the need for additional devices was developed. We propose a detection uses digital image processing analyze images of blister detect presence pills. The circular Hough transform as feature extraction technique and is therefore primarily useful pills with round shape. This composed two steps. First, registration full storing reference values local database. Second, classification taken remaining similar blisters, determine actual number untaken In 100% gray blisters or transparent cover were successfully detected. counting partially opened 95.2% 95.1% detected while 88.2% 80.8% cover. proposed provides promising results However, further improved, particular non-oval shapes.