TechaScope

TechaScope is a vision assistant that uses computer vision and OCR to help recognize objects and read out any text it finds.
It combines YOLOv4-tiny object detection, Tesseract OCR, and Text-to-Speech (TTS) in a simple web app.

---

How it works

• The camera feed is accessed directly in the browser.
• When you press CAPTURE (or wait for auto-capture), a snapshot is sent to the backend.
• The backend (Flask + OpenCV) runs:
  • Object detection with YOLOv4-tiny
  • Text detection with Tesseract OCR
  • Builds a combined message (object + detected text)
• The browser then displays the results and can speak them aloud using built-in text-to-speech.

---

Features

• 🎥 Live camera input with capture button
• 🔁 Auto-capture mode every few seconds (can be toggled in code)
• 🔊 Text-to-Speech (TTS) to read detected objects and text aloud
• ✋ Manual control — you can choose when to trigger speech
• 🖼 Annotated image preview showing bounding boxes and recognized text
• 📖 OCR words are filtered with a dictionary to reduce noise

---

Running the program

1. Start the backend server:

   py app.py

   The server runs on http://127.0.0.1:5000.

2. Open the frontend:

   • In your browser, go to http://127.0.0.1:5000
   • Allow camera access
   • Press CAPTURE or let it auto-capture

3. See results on the page:

   • Detected object name
   • Item name from OCR text
   • Full speech text (also spoken aloud if enabled)

---

Notes

• Auto-read can be turned on/off in the frontend JavaScript.
• Confidence thresholds and capture intervals can be adjusted in the code.
• Model weights (yolov4-tiny.weights) are required locally but ignored in GitHub.

---

Authors

Shan Truong