Smart Camera for Object Detection, Distance and Relative Speed Estimation

Abstract

The aim of this project is to detect moving objects, estimate their distance from the camera and return these values to the user in a dynamic setup. The values returned are name of the object, accuracy in detection, x and y coordinates which will enable in calculating parameters like speed and distance of the object. Object detection is done using Deep learning using Neural Network. and is implemented using OpenCV and Python’s built in dnn (deep neural network) function. The MobileNet SSD used here is a Caffe version of the original TensorFlow implementation. This project combines object detection with distance estimation and speed to return the location for the required objects.