Object detection with YOLO-V2 and voice alert for visually impaired is a Python project implemented with YOLO pre-trained model. This uses webcamera, it can used either externally or through laptop camera for demonstration. The system generates the voice alert and movement in direction such as ‘Right’ or ‘Left’ is also specified on the output. The object detected for the demonstration purpose on the given demo video is Person and Mobile phone. Whenever the objects moves, it gives voice alert to user.
How Object detection can be implemented with YOLO?
YOLO is the pre-trained model, which can detect more than 9000 classes of objects. The pre-trained weight files and class name files are loaded to the project directory. Opencv is the library used for YOLO object detection. More information of YOLO object detection can be viewed at https://opencv-tutorial.readthedocs.io/en/latest/yolo/yolo.html
How to generate Voice alert for Object detection project?
Voice alert can be integrated to our application through Python code. GTTS (Google Text to Speech) is the library used generally for conversion of detected object as text value to Speech data. This is very fast in conversion and more reliable to use.
How to get Architecture diagram for Object detection project?
The following figure shows the architecture diagram of the object detection project. This diagram shows all modules implemented on this work. Input for the project is Video stream converted to image frames for object detection detection. Output of the project is detected object is shown with bounding boxes and voice output.
Demo Video of Object Detection project
Related Object detection project are as follows