Real-Time Object Detection and Distance Measurement

AccelATHON 3.0

Team Name: Cipher

Problem Statement

Real World Distance measurement by detecting and identifying the Object using WebCam. A prototype that can help blind people navigate smoothly.

Algorithm used: YoloV4:Yolo i.e., You only look once. Dataset uses: COCO i.e., Common Objects in Context

Future Goal: Implementing the project on Hardware kit, i.e., ESP32 CAM.

Requirements

pip install -r requirements.txt

Code Flow

• Import OpenCV Module and create a function to detect objects.
• Create two more functions to calculate Distance and Focal length
• Create an instance of DetectionModel and pass reference images to the detected object.
• Using cv2, capture the real-time Video using WebCam and measure distance in inches.

Screenshots

Presentation

Canva Presentation- Team Cipher

Future Aspects- Applications

• Self Driving Cars
• Vision to Blind
• Argumented Reality

Conclusion

• In the given problem statement, Objects were detected successfully.
• To detect the Objects successfully YOLO dataset came in handy.
• Implemented a Prototype for the problem statement i.e., to find the distance.
• In future this problem statement code will be dumped on Raspberry PI or ESP-32

Authors

• @TRJ_0751