Anemia is a common blood disorder that reduces the oxygen-carrying capacity of blood and may lead to fatigue, weakness, and other serious complications. Early detection is essential for timely treatment. This project presents a non-invasive, AI-based Anemia Detection System using Eye Conjunctiva Analysis implemented on Raspberry Pi 3 B+ to identify anemic and non-anemic individuals.

The system uses a Pi Camera to capture high-resolution images of the lower eyelid (conjunctiva), where visible color variations can indicate signs of anemia. The captured image is processed using image processing and artificial intelligence techniques. The system converts the image from RGB to HSV (Hue, Saturation, Value) color space to effectively extract color features. The HSV method helps isolate redness intensity by separating color components from brightness levels, making it easier to detect pale conjunctiva associated with anemia.

The Raspberry Pi 3 B+ performs image acquisition, preprocessing (noise reduction and color normalization), region of interest (ROI) extraction, HSV-based feature extraction, and classification using a trained machine learning model. Based on the extracted HSV features, the system classifies the individual as either Anemic or Non-Anemic.

The result is displayed on an LCD screen showing the status (Anemic / Non-Anemic). If anemia is detected, a buzzer alert is activated to immediately notify the user.

This system provides a cost-effective, portable, and non-invasive screening solution suitable for rural and remote healthcare environments where laboratory testing facilities are limited. The project demonstrates how embedded systems combined with AI and HSV-based image analysis can assist in early medical screening.

    The main objectives of this project are:

• To capture conjunctiva images using Pi Camera.
• To preprocess images (noise removal and normalization).
• To convert RGB images to HSV color space.
• To extract color intensity features from the conjunctiva region.
• To train a classification model for anemia detection.
• To display results on LCD.
• To provide buzzer alert if anemia is detected.
• To develop a portable and cost-effective screening system.

  The major building blocks of this project are:  

1. Adapter power supply.
2. Raspberry pi3 B+..
3. SD card.
4. Pi camera.
5. Buzzer.
6. LCD display.

  Software’s used:  

• Raspbian OS.

• Express SCH for Circuit design.
• Image processing.

 

video: