Page 324 - AI Computer 10

F. Long answer type questions.

1. Explain the concept of Computer Vision and its role in arti fi cial intelligence.
Ans. Computer Vision is a fi eld in AI that teaches machines to interpret and understand the visual world. It
deals with extracti ng informati on from images, including videos and 3D scans. Its applicati ons range
from manufacturing to healthcare, where machines mimic human visual percepti on.

2. How does Computer Vision contribute to the manufacturing industry, and what is the role of machine
vision?

Ans. In manufacturing, Computer Vision aids in automated inspecti ons and identi fying defects on producti on
lines. Machine vision, a closely allied fi eld, opti mizes operati onal processes by detecti ng irregular events
and inconsistencies.
3. Discuss the role of Computer Vision in Google Translate, highlighti ng the camera feature’s signifi cance.

Ans. Google Translate’s camera feature uti lizes Computer Vision to translate large blocks of text by capturing
images. Users can point their phone’s camera at foreign language text, and the app detects the source
language, providing translati ons without manual typing.

4. Explain the applicati ons of Computer Vision in the healthcare industry, focusing on medical imaging.
Ans. Medical imaging, a term associated with healthcare, uti lizes Computer Vision for visualizing the
body’s interiors. It plays a crucial role in diagnosing diseases, off ering doctors effi cient tools for clinical
examinati on and interventi on.

5. Explain the concept of Opti cal Character Recogniti on (OCR) in the context of Computer Vision, emphasizing
its applicati ons.

Ans. OCR, a technology within Computer Vision, automates encoding and transcripti on of digital documents.
It extracts informati on from documents and PDFs, enhancing the documentati on process, saving ti me,
and improving customer operati ons.

G. Competency-based questions.
1. Atharv wants to develop an image processing applicati on using Computer Vision. His senior developer
advised him to choose a CNN for the model instead of a traditi onal neural network. What could be the
reason for this advice?
a. CNNs have fewer parameters, making them more effi cient for image processing.
b. CNNs are simpler to develop than traditi onal neural networks.

c. CNNs do not require large amounts of training data.
d. CNNs cannot be used for tasks like object detecti on.
2. Aanya was studying about CNNs. But she is not clear about the role of convoluti onal layer in a CNN. Help
her by describing the role of a convoluti onal layer.
a. Reduce the dimensionality of input data

b. Extract spati al features by applying fi lters to input images
c. Convert images into numerical vectors for classifi cati on

d. Flatt en the image for further processing
3. Antriksh was developing an image processing applicati on that uses computer vision and convoluti ons for
extracti ng features from an image. He chose an image input matrix and a suitable kernel matrix for the
convoluti ons. But, he was not able to get desired output. What could be the reason for this error?

a. He forgot to fl ip the input matrix.
b. He forgot to create the output matrix.

190
190