cvpr maths
Budget: £10 – £20 GBP
i want this questions answered:
They asked you to design a computer vision system for the automated detection of tools on workshop
benches. Overhead cameras have been installed to provide images of benches in the
workshops. All workshop benches have the same colour, and the system should recognise
4 different tools as shown in Figure 1(a). The workshop is open 24 hours and there is a
mixture of natural and artificial illumination. The system should recognise 4 individual
tools shown in Fig. 1 (combination pliers, adjustable wrench, screwdriver “-“, and
screwdriver “+”).
///////
1st : Draw a block diagram of the descriptor extraction part of your proposed recognition
system, briefly explaining the functionality of each block, and name one technique
that implements each functionality. You are free to use any techniques covered in
lectures, but your system should at least achieve foreground/background segmentation, find individual tools in the field of view, and extract a descriptor of each tool
to support recognition of individual tools.
2nd:Suggest a robust model-based approach to determine if a pixel belongs to a foreground (i.e. a tool) or background (i.e. bench). How many parameters does your
model have? What is the dimension of the parameter space that your model uses?
You should give the details about the equations for building model parameters and
describe the parameters that can be adjusted (you do not have to describe how to
find the values of tuneable parameters).
3rd: Explain how to use your model to decide if a pixel belongs to the background or
foreground. Write an appropriate equation supporting the foreground/background
decision.
They asked you to design a computer vision system for the automated detection of tools on workshop
benches. Overhead cameras have been installed to provide images of benches in the
workshops. All workshop benches have the same colour, and the system should recognise
4 different tools as shown in Figure 1(a). The workshop is open 24 hours and there is a
mixture of natural and artificial illumination. The system should recognise 4 individual
tools shown in Fig. 1 (combination pliers, adjustable wrench, screwdriver “-“, and
screwdriver “+”).
///////
1st : Draw a block diagram of the descriptor extraction part of your proposed recognition
system, briefly explaining the functionality of each block, and name one technique
that implements each functionality. You are free to use any techniques covered in
lectures, but your system should at least achieve foreground/background segmentation, find individual tools in the field of view, and extract a descriptor of each tool
to support recognition of individual tools.
2nd:Suggest a robust model-based approach to determine if a pixel belongs to a foreground (i.e. a tool) or background (i.e. bench). How many parameters does your
model have? What is the dimension of the parameter space that your model uses?
You should give the details about the equations for building model parameters and
describe the parameters that can be adjusted (you do not have to describe how to
find the values of tuneable parameters).
3rd: Explain how to use your model to decide if a pixel belongs to the background or
foreground. Write an appropriate equation supporting the foreground/background
decision.