wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Content Section Worksheet Questions (Grade 1 adapted MCQs)

Total questions: 100

Worksheet time: 50mins

Name
Class
Date
1.

A face detection system using Viola–Jones gives many false positives in cluttered backgrounds. Which change logically reduces this problem?

a)

Increase image resolution

b)

Add more stages to the cascade

c)

Remove AdaBoost

d)

Use grayscale images

2.

YOLO detects objects faster than R-CNN. The main logical reason is that YOLO:

a)

Uses fewer convolution layers

b)

Avoids region proposals

c)

Uses handcrafted features

d)

Works only on videos

3.

If YOLO fails to detect small objects, the most logical improvement is to:

a)

Increase learning rate

b)

Reduce number of classes

c)

Increase grid resolution or use multi-scale detection

d)

Convert images to grayscale

4.

A CNN trained on daytime images performs poorly at night. The logical cause is:

a)

Overfitting to illumination conditions

b)

Poor optimizer

c)

Incorrect labels

d)

Insufficient epochs

5.

Transfer learning improves performance when training data is limited because:

a)

It avoids convolution

b)

Pretrained models capture generic visual features

c)

It reduces network depth

d)

It increases batch size

6.

Optical flow estimation becomes unreliable in textureless regions because:

a)

Motion is zero

b)

Brightness constancy fails

c)

No intensity variation exists to track motion

d)

Frames are identical

7.

Horn–Schunck optical flow produces smoother motion fields than Lucas–Kanade because it:

a)

Uses local windows

b)

Enforces global smoothness constraint

c)

Ignores brightness constancy

d)

Uses feature matching

8.

If fast-moving objects appear blurred, optical flow accuracy decreases mainly because:

a)

Frame rate is low

b)

Background is static

c)

Motion is uniform

d)

Image resolution is high

9.

A GMM background model fails when lighting changes suddenly. The logical fix is to:

a)

Increase number of Gaussians

b)

Reduce learning rate adaptation time

c)

Remove background subtraction

d)

Convert to binary images

10.

Multiple Gaussians per pixel are required because:

a)

Foreground objects are static

b)

Background may have repetitive variations

c)

Noise is eliminated

d)

Color images are used

11.

If a pixel consistently appears in foreground for a long time, GMM will:

a)

Always mark it as foreground

b)

Ignore it permanently

c)

Adapt and include it in background

d)

Remove the pixel

12.

Structure from Motion requires multiple images because depth cannot be inferred from:

a)

Stereo images

b)

Motion cues

c)

A single static image

d)

Feature correspondences

13.

If camera motion is incorrectly estimated in SfM, the reconstructed 3D structure will be:

a)

More accurate

b)

Unaffected

c)

Distorted

d)

Faster to compute

14.

Feature matching errors in SfM mostly affect:

a)

Color consistency

b)

3D point accuracy

c)

Image resolution

d)

Compression

15.

Motion estimation is preferred over optical flow in video compression because it:

a)

Is pixel-accurate

b)

Is computationally simpler

c)

Captures object semantics

d)

Uses deep learning

16.

Block-based motion estimation fails in regions with uniform intensity because:

a)

No blocks exist

b)

Matching cost becomes ambiguous

c)

Motion is large

d)

Noise is high

17.

If Mean Squared Error is minimized in block matching, the algorithm favors:

a)

Sharp edges

b)

Maximum motion

c)

Closest pixel similarity

d)

Global motion

18.

Image analysis focuses on feature extraction because raw pixels:

a)

Are noisy

b)

Are redundant and high-dimensional

c)

Have low resolution

d)

Cannot be processed

19.

Texture features are better than color features when:

a)

Objects have similar colors

b)

Lighting is uniform

c)

Images are grayscale

d)

Resolution is high

20.

Stereo vision depth estimation becomes inaccurate when baseline is too small because:

a)

Disparity becomes negligible

b)

Images overlap

c)

Cameras misalign

d)

Epipolar lines vanish

21.

Large disparity corresponds to closer objects because:

a)

Camera motion increases

b)

Parallax effect is stronger

c)

Images blur

d)

Resolution increases

22.

Epipolar constraint improves stereo matching efficiency by:

a)

Reducing image size

b)

Limiting search to a line

c)

Increasing disparity

d)

Removing noise

23.

Motion models are required in animation to ensure:

a)

Faster rendering

b)

Realistic temporal consistency

c)

Higher resolution

d)

Color correction

24.

Rigid motion models fail for human motion because:

a)

Humans move slowly

b)

Body parts deform independently

c)

Images are noisy

d)

Motion is linear

25.

Affine motion models are preferred over simple translation because they can handle:

a)

Noise

b)

Scaling and rotation

c)

Non-rigid deformation

d)

Illumination change

26.

A pattern recognition system fails mainly due to poor:

a)

Display resolution

b)

Feature selection

c)

Camera quality

d)

Storage capacity

27.

Compact feature vectors are preferred because they:

a)

Increase accuracy always

b)

Reduce computation and overfitting

c)

Improve brightness

d)

Increase dimensions

28.

Occlusion is a major challenge in object recognition because:

a)

Objects disappear

b)

Partial information is available

c)

Lighting changes

d)

Background is uniform

29.

Normalization improves recognition accuracy by:

a)

Increasing dataset size

b)

Reducing irrelevant variations

c)

Enhancing edges

d)

Increasing contrast

30.

Deep learning outperforms traditional methods mainly because it:

a)

Uses rules

b)

Learns hierarchical features

c)

Requires no data

d)

Avoids training

31.

Overfitting occurs when a model performs well on training but poorly on testing due to:

a)

Under-training

b)

Excessive generalization

c)

Memorization of training data

d)

Poor labels

32.

Data augmentation improves generalization by:

a)

Reducing model size

b)

Introducing controlled variability

c)

Removing noise

d)

Increasing epochs

33.

Object recognition accuracy improves when motion cues are added because:

a)

Color improves

b)

Temporal consistency is exploited

c)

Resolution increases

d)

Noise is removed

34.

A real-time system sacrifices some accuracy mainly to achieve:

a)

Better training

b)

Lower latency

c)

Higher resolution

d)

Larger models

35.

Why is GMM preferred over simple thresholding in dynamic scenes?

a)

Faster computation

b)

Adaptive background modeling

c)

Less memory

d)

Simpler math

36.

Stereo vision is superior to monocular vision for depth because it:

a)

Uses learning

b)

Uses geometry

c)

Needs less data

d)

Avoids calibration

37.

If correspondence between stereo images is incorrect, depth estimation will be:

a)

Accurate

b)

Unchanged

c)

Erroneous

d)

Faster

38.

Motion estimation improves tracking because it:

a)

Reduces noise

b)

Predicts object position

c)

Enhances color

d)

Increases resolution

39.

In animation, non-rigid motion is necessary to model:

a)

Vehicles

b)

Buildings

c)

Cloth and human skin

d)

Cameras

40.

Logical advantage of CNNs over traditional classifiers is their ability to:

a)

Use fewer layers

b)

Automatically learn spatial features

c)

Avoid optimization

d)

Eliminate training

41.

If training and test data distributions differ, performance drops due to:

a)

Overfitting

b)

Dataset bias

c)

Underflow

d)

Noise removal

42.

Combining stereo vision and motion estimation improves depth accuracy because:

a)

Both reduce noise

b)

Multiple cues reinforce estimation

c)

Computation increases

d)

Resolution improves

43.

Feature redundancy negatively affects classifiers because it:

a)

Improves accuracy

b)

Increases computation without new information

c)

Reduces data

d)

Removes noise

44.

Why are CNNs robust to small translations?

a)

High learning rate

b)

Pooling operations

c)

Fully connected layers

d)

Softmax

45.

Logical drawback of very deep networks is:

a)

Better accuracy

b)

Vanishing gradients

c)

Less data usage

d)

Faster training

46.

Motion blur affects object recognition because it:

a)

Improves edges

b)

Removes high-frequency details

c)

Increases contrast

d)

Stabilizes motion

47.

Temporal consistency helps object recognition in video by:

a)

Ignoring frames

b)

Using previous frame information

c)

Reducing frame rate

d)

Removing motion

48.

A model trained only on frontal faces fails on side views due to:

a)

Low resolution

b)

Viewpoint variation

c)

Noise

d)

Occlusion

49.

Why is illumination invariance important in pattern recognition?

a)

Lighting never changes

b)

Objects appear different under different lighting

c)

Images are grayscale

d)

Cameras are fixed

50.

Integrating detection, recognition, and tracking improves systems because it:

a)

Increases complexity only

b)

Uses complementary information

c)

Reduces memory

d)

Eliminates errors

51.

In the Viola–Jones algorithm, integral images are primarily used to:

a)

Reduce false positives using cascades

b)

Compute Haar-like features in constant time

c)

Improve image resolution

d)

Classify images using CNNs

52.

Why is AdaBoost critical in Viola–Jones face detection?

a)

It enhances pixel contrast

b)

It selects and weights the most discriminative features

c)

It segments the image

d)

It performs motion tracking

53.

Which property of Haar-like features makes them suitable for real-time detection?

a)

Rotation invariance

b)

Low computational complexity

c)

High texture sensitivity

d)

Color dependence

54.

The cascade classifier in Viola–Jones improves performance mainly by:

a)

Using deep neural networks

b)

Rejecting negative regions early

c)

Increasing training data

d)

Applying frequency-domain filtering

55.

YOLO treats object detection as a:

a)

Classification problem

b)

Region proposal problem

c)

Regression problem

d)

Clustering problem

56.

Why does YOLO achieve faster detection compared to R-CNN-based methods?

a)

Uses handcrafted features

b)

Processes the entire image in a single forward pass

c)

Uses smaller datasets

d)

Avoids convolution operations

57.

In YOLO, grid cells are responsible for:

a)

Image enhancement

b)

Motion estimation

c)

Predicting bounding boxes and class probabilities

d)

Stereo correspondence

58.

A limitation of YOLO is its difficulty in detecting:

a)

Large objects

b)

Textured objects

c)

Small objects close together

d)

Static objects

59.

Which deep learning architecture is most commonly used for object recognition?

a)

RNN

b)

CNN

c)

GAN

d)

Autoencoder

60.

CNNs are effective for object recognition mainly because they:

a)

Use handcrafted features

b)

Exploit spatial hierarchy using convolution

c)

Reduce image size only

d)

Perform unsupervised clustering

61.

Which layer in a CNN provides translation invariance?

a)

Convolution layer

b)

Fully connected layer

c)

Pooling layer

d)

Softmax layer

62.

Transfer learning is useful in object recognition because it:

a)

Eliminates training

b)

Reuses learned features from large datasets

c)

Reduces image noise

d)

Improves optical flow accuracy

63.

Optical flow estimates:

a)

Object boundaries

b)

Pixel-wise motion between frames

c)

Image depth

d)

Color consistency

64.

The brightness constancy assumption in optical flow states that:

a)

Brightness changes over time

b)

Pixel intensity remains constant along motion

c)

Motion is linear

d)

Objects do not move

65.

Gaussian Mixture Models (GMM) are commonly used in vision for:

a)

Edge detection

b)

Background subtraction

c)

Stereo matching

d)

Object classification

66.

In GMM-based background modeling, each pixel is represented as:

a)

Single Gaussian

b)

Binary value

c)

Mixture of Gaussians

d)

Histogram

67.

Why are multiple Gaussians used per pixel in GMM?

a)

To reduce noise

b)

To model complex background variations

c)

To enhance contrast

d)

To increase frame rate

68.

Foreground objects in GMM are detected when pixel values:

a)

Match dominant Gaussians

b)

Do not fit background distributions

c)

Have high brightness

d)

Are stationary

69.

Structure from Motion (SfM) aims to recover:

a)

Object color

b)

Camera calibration only

c)

3D structure and camera motion

d)

Optical flow vectors

70.

SfM primarily relies on:

a)

Single image analysis

b)

Multiple views of a scene

c)

Stereo pairs only

d)

Infrared images

71.

Which feature is commonly used in SfM pipelines?

a)

Haar features

b)

SIFT / ORB

c)

Histogram equalization

d)

Thresholding

72.

Motion estimation differs from optical flow because it:

a)

Is pixel-based

b)

Estimates block or global motion

c)

Uses color models

d)

Detects objects

73.

Block-based motion estimation is widely used in:

a)

Object recognition

b)

Image enhancement

c)

Video compression

d)

Stereo vision

74.

Which metric is commonly minimized in block matching algorithms?

a)

Entropy

b)

Mean Squared Error

c)

Precision

d)

Recall

75.

Image analysis in pattern recognition primarily focuses on:

a)

Image compression

b)

Extracting meaningful features

c)

Reducing resolution

d)

Color correction

76.

Which feature is most suitable for texture analysis?

a)

Color histogram

b)

Edge orientation

c)

GLCM features

d)

Bounding box size

77.

Stereo vision estimates depth using:

a)

Motion vectors

b)

Disparity between image pairs

c)

Optical flow

d)

GMM

78.

Greater disparity in stereo vision indicates:

a)

Object is farther

b)

Object is closer

c)

No depth

d)

Flat surface

79.

The epipolar constraint reduces correspondence search to:

a)

Entire image

b)

Horizontal line

c)

Vertical line

d)

Epipolar line

80.

Which factor most affects stereo matching accuracy?

a)

Image brightness

b)

Camera baseline

c)

Color depth

d)

Compression ratio

81.

Motion models for image animation describe:

a)

Pixel colors

b)

Object movement over time

c)

Image resolution changes

d)

Noise patterns

82.

Rigid motion assumes that:

a)

Object shape changes

b)

Object deforms

c)

Distances between points remain constant

d)

Motion is random

83.

Which is an example of non-rigid motion?

a)

Rotating wheel

b)

Translating car

c)

Walking human

d)

Moving box

84.

Affine motion model can represent:

a)

Only translation

b)

Translation and rotation only

c)

Translation, rotation, scaling, shear

d)

Non-linear deformation

85.

Why are motion models important in animation?

a)

Reduce storage

b)

Produce realistic movement

c)

Enhance brightness

d)

Remove noise

86.

Pattern recognition systems typically consist of:

a)

Sensor, feature extractor, classifier

b)

Encoder, decoder, filter

c)

Camera, display, memory

d)

Histogram, threshold, output

87.

Feature vectors should ideally be:

a)

High dimensional

b)

Redundant

c)

Discriminative and compact

d)

Random

88.

Which classifier is commonly used in object recognition tasks?

a)

KNN

b)

SVM

c)

Decision Tree

d)

All of the above

89.

Why is normalization important in pattern recognition?

a)

Increase image size

b)

Reduce variation due to scale and illumination

c)

Improve color

d)

Detect motion

90.

Which challenge is common in real-world object recognition?

a)

Uniform background

b)

Controlled lighting

c)

Occlusion

d)

Fixed viewpoints

91.

Deep learning reduces the need for manual feature design because it:

a)

Uses predefined rules

b)

Learns features automatically

c)

Uses templates

d)

Ignores data

92.

Overfitting in deep object recognition models occurs when:

a)

Model generalizes well

b)

Training loss is high

c)

Model memorizes training data

d)

Dataset is large

93.

In the context of motion estimation, which technique is primarily used to track moving objects across frames?

a)

Image segmentation

b)

Optical flow

c)

Color histogram

d)

Feature extraction

94.

What is the primary advantage of using convolutional layers in a CNN for image processing?

a)

They increase image resolution

b)

They simplify the model architecture

c)

They reduce the number of parameters

d)

They enhance color accuracy

95.

In stereo vision, the disparity map is crucial for:

a)

Identifying object boundaries

b)

Estimating depth information

c)

Enhancing image contrast

d)

Reducing noise in images

96.

What is the primary advantage of using convolutional neural networks (CNNs) for image classification?

a)

They require less data preprocessing

b)

They are easier to implement

c)

They are faster than traditional methods

d)

They automatically learn spatial hierarchies

97.

In the context of image segmentation, what does the term 'superpixel' refer to?

a)

A pixel that has been compressed

b)

A group of pixels with similar colors

c)

A pixel with enhanced brightness

d)

A pixel that is part of a larger object

98.

Which technique is commonly used to improve the robustness of feature matching in varying lighting conditions?

a)

Histogram equalization

b)

Edge detection

c)

Image blurring

d)

Color inversion

99.

What is the primary advantage of using Convolutional Neural Networks (CNNs) for image classification?

a)

They require less data preprocessing

b)

They can learn spatial hierarchies of features

c)

They eliminate the need for feature extraction

d)

They are faster than traditional algorithms

100.

In the context of image segmentation, what does the term 'pixel-wise classification' refer to?

a)

Classifying entire images

b)

Reducing image resolution

c)

Assigning a class label to each pixel

d)

Detecting edges in images