Object Detection & Tracking — MCQs | Digital Image Processing

1. What is the main goal of object detection in image processing?

(A) Compressing image data

(B) Identifying and locating objects in an image

(C) Enhancing image sharpness

(D) Removing noise from images

2. Which algorithm is widely used for real-time object detection?

(A) SIFT

(B) YOLO

(C) PCA

(D) Canny

3. Which of the following is a region-based object detection algorithm?

(A) SSD

(B) YOLO

(C) R-CNN

(D) K-means

4. Which metric is commonly used to evaluate the performance of an object detector?

(A) PSNR

(B) SSIM

(C) mAP

(D) MSE

5. What does IOU stand for in object detection?

(A) Input Overlap Utility

(B) Intersection Over Union

(C) Integrated Object Unit

(D) Image Output Unit

6. Which object tracking method uses a histogram of the object’s appearance?

(A) Kalman Filter

(B) Mean Shift

(C) Optical Flow

(D) Deep SORT

7. Which tracking method assumes linear motion and Gaussian noise?

(A) Particle Filter

(B) Mean Shift

(C) Kalman Filter

(D) Optical Flow

8. Which technique is best suited for multi-object tracking?

(A) SURF

(B) Deep SORT

(C) Canny Edge Detector

(D) GLCM

9. In object detection, what does Non-Maximum Suppression (NMS) do?

(A) Blurs overlapping regions

(B) Selects the highest scoring bounding box among overlapping boxes

(C) Rescales image sizes

(D) Suppresses image noise

10. Which model is known for its speed in real-time object detection?

(A) R-CNN

(B) YOLO

(C) VGG16

(D) AlexNet

11. Which algorithm is a single-shot detector?

(A) Faster R-CNN

(B) SSD

(C) HOG

(D) KLT Tracker

12. Which feature descriptor is commonly used with object tracking?

(A) HOG

(B) Gabor

(C) LBP

(D) GLCM

13. Which algorithm integrates detection and tracking using deep learning?

(A) RANSAC

(B) Deep SORT

(C) LBP

(D) Mean Shift

14. What is the role of anchor boxes in object detection?

(A) To normalize image dimensions

(B) To provide reference shapes for object prediction

(C) To store pixel values

(D) To enhance contrast

15. Which of the following is NOT a challenge in object tracking?

(A) Occlusion

(B) Illumination changes

(C) Static background

(D) Scale variation

16. Which object detection algorithm is two-stage?

(A) YOLO

(B) SSD

(C) R-CNN

(D) MobileNet

17. What is the main input to an object tracker?

(A) Object histogram

(B) Detected bounding boxes

(C) Audio signals

(D) Optical flow maps

18. Which of the following is used to estimate motion in video frames?

(A) GLCM

(B) Optical Flow

(C) DFT

(D) LBP

19. Which object detection method uses selective search?

(A) YOLO

(B) SSD

(C) R-CNN

(D) Deep SORT

20. Which tracking algorithm is based on sampling and weights?

(A) Kalman Filter

(B) Mean Shift

(C) Particle Filter

(D) CNN

21. Which method is commonly used to reduce multiple detections of the same object?

(A) K-Means Clustering

(B) Gaussian Blurring

(C) Non-Maximum Suppression

(D) Thresholding

22. Which step is performed first in object detection?

(A) Bounding box regression

(B) Classification

(C) Feature extraction

(D) Post-processing

23. Which of the following uses deep learning for feature extraction?

(A) Haar Cascades

(B) YOLO

(C) Mean Shift

(D) Optical Flow

24. What is the purpose of the confidence score in detection?

(A) Measure of frame rate

(B) Indicates likelihood of object presence

(C) Counts number of pixels

(D) Measures distance to the camera

25. What does “real-time” mean in the context of object detection?

(A) Detection after video ends

(B) Detection within a minute

(C) Detection at or near camera capture speed

(D) Detection after preprocessing

26. Which of the following methods is NOT typically used in object tracking?

(A) Kalman Filter

(B) Particle Filter

(C) YOLO

(D) Optical Flow

27. Which deep learning model is widely used for feature extraction in detection?

(A) VGG

(B) RANSAC

(C) SURF

(D) DCT

28. What causes “drift” in object tracking?

(A) High frame rate

(B) Incorrect motion estimation over time

(C) Low resolution

(D) Poor lighting

29. Which layer in a neural network performs detection in YOLO?

(A) Input Layer

(B) Fully Connected Layer

(C) Output Layer

(D) Detection Layer

30. What makes SSD faster than R-CNN?

(A) Uses pre-trained filters

(B) Single-stage detection

(C) Works on grayscale images

(D) Reduces the number of features

31. What is the main output of an object detection model?

(A) Segmented regions

(B) Binary mask

(C) Bounding boxes with class labels

(D) Histograms

32. Which feature is used in the HOG descriptor?

(A) Color

(B) Texture

(C) Gradient orientation

(D) Intensity levels

33. Which is a disadvantage of deep learning–based detection methods?

(A) Slow inference

(B) Requires large labeled datasets

(C) Doesn’t work on videos

(D) No accuracy

34. Which detector is trained to predict multiple classes and boxes?

(A) PCA

(B) YOLO

(C) Gabor

(D) DCT

35. What causes tracking failure in Mean Shift?

(A) Uniform histogram

(B) Complex background

(C) Fast motion

(D) Noisy images

36. Which loss function is commonly used in object detection?

(A) Binary Cross-Entropy

(B) Mean Squared Error

(C) Smooth L1

(D) Euclidean Loss

37. Which tracker adapts better to non-linear motion?

(A) Kalman Filter

(B) Optical Flow

(C) Particle Filter

(D) CNN

38. Which framework combines object detection and tracking in one system?

(A) CNN

(B) Deep SORT

(C) DFT

(D) PCA

39. What does the “Faster” in Faster R-CNN refer to?

(A) Faster training

(B) Use of GPU

(C) Region Proposal Network

(D) Small image input

40. Which object detection model is most suitable for mobile applications?

(A) YOLOv3

(B) MobileNet SSD

(C) R-CNN

(D) Fast R-CNN

41. Which part of YOLO predicts object locations?

(A) Region Proposal Network

(B) Bounding Box Regressor

(C) Grid cells

(D) Confidence map

42. Which tracker uses prior and observation models?

(A) Kalman Filter

(B) Particle Filter

(C) Mean Shift

(D) KLT

43. Which method uses pyramids in object detection?

(A) R-CNN

(B) HOG

(C) YOLO

(D) Deep SORT

44. Which issue is common in long-term tracking?

(A) Frame skipping

(B) Background modeling

(C) Occlusion

(D) Frame resizing

45. Which detection model uses regression directly to predict boxes?

(A) R-CNN

(B) YOLO

(C) Deep SORT

(D) SIFT

46. Which detector divides the image into grids?

(A) R-CNN

(B) SSD

(C) YOLO

(D) GLCM

47. What does the Kalman Filter estimate in tracking?

(A) Color distribution

(B) Motion and state of object

(C) Shape of object

(D) Image gradients

48. What type of learning is used in most modern object detection systems?

(A) Supervised learning

(B) Unsupervised learning

(C) Reinforcement learning

(D) Semi-supervised learning

49. Which algorithm is best for low-computation environments?

(A) Faster R-CNN

(B) YOLOv5

(C) SSD with MobileNet

(D) Deep SORT