April
1999 QUESTION 4 Total Marks: 20 Marks |
Click here to access other
questions
GRADE A
|
| (a) | The difference between low and high level
vision is in the kind of knowledge they use. Describe high and low level vision with regards to the kind of knowledge used. |
[4] |
| In low level vision,
it uses bottom up and parallel processing. The kind of knowledge is used to determine the
validity of the information (such as is it accurate and complete). In high level vision, it uses goal achieve and spatially focus process. It makes use of the information to solve complex problem such as navigation.
|
||
| (b) | Extracting information from an image is a complex task. Even a task like determining if a point is within an object, which is simple to human, is not easily achieved by computers. Ray-intersection is a technique that can achieve this. Briefly describe this technique. | [4] |
| A ray is drawn from
the position in question to the edge of the image. The intersection between the ray and
boundary is count. If the count is even, then the position is outside. The image otherwise
it is inside the image.
|
||
| (c) | Briefly describe how patterns can be recognized using feature analysis. | [4] |
| If assume that the
input object consists of distinct and separable parts called features. Pattern recognition
is achieve by counting the presence or absence of features and compare with a fabrication
of features associated with the object.
|
||
| (d) | One way of performing pattern recognition is by using template matching. Using appropriate diagrams and examples, describe two problems of this approach. | [8] |
|