Butterfly Project – Camera Recognition
A mobile app that allows users to record rare butterflies using the smartphone camera. By taking a photo, the app identifies the butterfly species, automatically fetches an image from a database, and shows it for confirmation.
Key Issues (Challenges)
- Image Recognition: Reliable butterfly detection and classification (AI model training)
- Image Database: Needs large, high-quality reference datasets for accurate recognition
- Accuracy & Validation: Risk of misclassification between visually similar species
- User Experience: Clear confirmation screen, easy re-try if recognition fails
- Performance: On-device vs. cloud recognition (speed, data privacy, battery usage)
Stakeholders
- Users: Nature enthusiasts, schools, conservationists
- App Developers: Implement camera + AI recognition pipeline
- AI Model Providers: Pretrained or custom models
- Research & Conservation Organizations: Receive validated observation data
Purpose / Benefits
- Make butterfly recognition accessible to everyone
- Reduce dependency on prior knowledge
- Encourage citizen science participation
- Support conservation efforts with reliable image-based data
Empathy Map
Says: “I want to identify this butterfly quickly.”, “I’m not sure if this is a rare species.”, “It should be easy to use, even in the field.”, “I hope the app works even without internet.”, “I want to share my findings with others.”
Thinks: “Will the recognition be accurate enough?”, “Is my data and location private?”, “This could help science and conservation.”, “Will the app give me suggestions if it’s unsure?”, “I don’t want to waste time if the recognition fails.”
Does: Uses the camera or video mode, adjusts angle/lighting, waits for recognition, checks confirmation, shares results, uploads offline if needed.
Feels: Excited when identifying a rare butterfly, curious to learn more, frustrated if recognition is wrong or slow, motivated to contribute, proud when validated.
Target Customers
- Nature enthusiasts & hobby photographers
- Students & schools (biology classes, field trips)
- Researchers & conservation organizations
- Citizen scientists / eco-volunteers
Top 10 Critical Issues
- Accuracy under different conditions
- Handling of rare/similar species
- Speed of recognition
- Data privacy & GDPR
- Offline mode vs internet dependency
- Quality/licensing of image database
- Battery & performance impact
- Ease of use for schools/kids
- Trust in results (confidence scores)
- Integration with research platforms
Pain / Gain
Pain: Wrong or slow recognition, limited offline functionality, privacy concerns, battery drain, lack of trust.
Gain: Fast and accessible identification, awareness of biodiversity, contribution to conservation, pride in helping science, strong community of observers.
Personas
Anna (32, Nature Enthusiast) – Wants quick IDs and sharing, needs easy recognition and offline use, frustrated by slow or inaccurate results.
Leon (15, Student Explorer) – Uses app in school projects, needs simple interface and educational features, frustrated if too complex or unexplained.
Dr. Fischer (45, Researcher) – Needs accurate IDs, confidence scores, research integration, GDPR compliance; frustrated by low quality data or too much gamification.
Disney Method
Dreamer: Global app for instant ID, worldwide database, school quizzes, future expansion to other animals.
Realist: AI image recognition models, mobile app workflow, offline caching, step-by-step rollout.
Critic: Risks of misidentification, privacy issues, dataset costs, connectivity challenges, user motivation.
Controller: Balance gamification and science, strong privacy rules, regular updates, partnerships with schools/NGOs, monitor feedback.