A company wants to design a system to detect different types of objects in a conveyor belt and get their position to be picked up by a robot. Fig. 1 and Fig. 2 show two different types of products to be detected. The conveyor belt has a height of 50cm with respect to the ground, a width of 80cm and its velocity is 0,5m/s. The questions are: a. The conveyor belt has a video camera located at 1m of height, in a zenithal position and perpendicular to it. Which is the lens angle of the camera to focus the camera in the width of the conveyor belt? Which is the number of pixels of the cameras to detect with a precision of 0,5mm ? If the camera takes images each 40ms, will the system be able to detect the objects with a precision of 0,5mm ? b. How will the system be able to detect the bottles of the conveyor belt of Fig.1? Which are the best features and computer vision techniques to be used? If the bottles are partially occluded, which features can be used? c. How will the system be able to detect the boxes with labels of the conveyor belt of Fig.2? Which are the best features and computer vision techniques to be used? If the boxes are partially overlapped, how can the system detect them