Reference no: EM132389212
Question 1 -
We first suppose we use one type of lens + camera bundle, whose focal length f is set to 2.8mm, and the pixel size tu is set to 6um, such as the OmniVision OV7962 camera module. We assume a resolution of 640 (W) x 480 (H) pixels.
What is the associated value for v you obtain?
To answer this question,
- You will first formulate the complete optimization problem, and justify all constraints put on the variables of the system: which one are fixed? Continuous? What is their bounds? etc.
- Then you'll use your preferred solver to solve this problem, e.g. scipy.optimize.minimize with the SLSQP method, PyOpt or embedded solvers from Matlab or other approach.
Question 2 -
The cost of elements in a car has a significant economic impact. A few dollars difference is enough to win competition. Therefore, we have to optimize systems for both efficiency and price.
We consider the following references, from DigiKey.
OV07962-E62A, f = 2.8mm, tu = 6 um, unit price = $7,191 / 1000 units
OV07690-R202A, f = 2.8mm, tu = 1.75 um, unit price = $4,312 / 1000 units
OVM9724-RYDA-ND, f = 1.66mm, tu = 1.4 um, unit price = $12,718 / 2500 units
For all modules, we assume a resolution of 640x480 pixels. Note that we need 2 modules for each car. For each car, we consider that it is worth it to spend $1 more if we can improve v (for the nearby obstacle with a height of 165cm) by at least 2 pixels.
Resolve again the optimization problem from Q1, which is the best bundle to pick? What is the associated value for v you obtain? How much does it cost in total for one car? Please show your intermediate steps for partial credits.
Note - Submit your answers as a pdf. Submit the optimization script (.py/.m or other format). Submit a readme file (readme.txt), describing how to run the script.
Attachment:- Assignment File.rar