Voice controlled Camera Assisted Pick and Place Robot Using Raspberry Pi

Abstract

Modern monitoring systems or manufacturing machines have a major drawback as they depend on human operators who can easily get distracted or make mistakes, so a system is needed that can constantly monitor the desired area and make decisions while identifying a pre-trained object. Tracking objects with a camera is critical in any automated monitoring and tracking system. The main goal of this paper is to design and implement a robot that can distinguish objects based on their features, such as color and size, and based on artificial intelligence and image processing algorithms.  The robot will analyze the video stream to detect the colored object, and specify its location inside the video frame. Using the detected position, the raspberry pi will decide the rotation direction whether it is to the right to the left, or forward until it reaches the object, grabs it and puts it in the robot's pocket.  The main controlling unit of the system is the Raspberry Pi, the robot is equipped with a Wi-Fi modem to communicate with the mobile application, which is used to control the robot in two modes: manual mode, where the user can point the robot in any direction either by pressing function button or through voice commands. The second mode is the Automatic mode, where the user can ask the robot to detect an object according to a set of characteristics and grab it without any human intervention and based on a novel digital image processing object-tracking algorithm, the accuracy in voice command mode has reached 95%.