1. Which filtering technique attempts to reverse the degradation process of an image?
(A) Median filtering
(B) Inverse filtering
(C) Gaussian filtering
(D) Laplacian filtering
2. Which filter minimizes the mean square error between estimated and original images?
(A) Gaussian filter
(B) Inverse filter
(C) Wiener filter
(D) Median filter
3. Inverse filtering is most effective when:
(A) The noise level is very high
(B) The degradation function is known and noise is minimal
(C) The original image is blurred by Gaussian noise
(D) The filter size is increased
4. Wiener filtering requires knowledge of:
(A) The degradation function only
(B) Only the noise characteristics
(C) Both degradation function and noise statistics
(D) Image resolution
5. Which of the following is a major drawback of inverse filtering?
(A) It is slow to compute
(B) It cannot handle blurring
(C) It amplifies noise
(D) It reduces resolution
6. Wiener filtering is typically applied in which domain?
(A) Spatial domain
(B) Time domain
(C) Frequency domain
(D) Histogram domain
7. In inverse filtering, division by very small values of the degradation function leads to:
(A) Better restoration
(B) Noise amplification
(C) Image sharpening
(D) Edge smoothing
8. The Wiener filter is optimal in the sense of:
(A) Minimizing variance
(B) Maximizing sharpness
(C) Minimizing mean square error
(D) Maximizing entropy
9. In practical applications, Wiener filtering is more stable than inverse filtering because:
(A) It uses edge detection
(B) It considers the signal and noise power spectra
(C) It avoids Fourier transform
(D) It performs local histogram equalization
10. When noise is negligible, the Wiener filter becomes similar to:
(A) Mean filter
(B) Gaussian filter
(C) Inverse filter
(D) Median filter
11. Which filtering method performs best when the noise characteristics are known?
(A) Median filter
(B) Inverse filter
(C) Wiener filter
(D) Bilateral filter
12. The inverse filter works poorly when:
(A) The blur kernel is large
(B) The image is smooth
(C) There is significant noise
(D) The image is high contrast
13. The main goal of inverse filtering is to:
(A) Enhance contrast
(B) Remove Gaussian noise
(C) Recover the original image from degraded version
(D) Sharpen edges
14. Wiener filtering adapts based on:
(A) Human vision model
(B) Histogram shape
(C) Local image statistics
(D) Edge density
15. The Wiener filter assumes:
(A) Additive Gaussian noise
(B) Multiplicative salt and pepper noise
(C) Only blur without noise
(D) No prior knowledge of noise
16. Which filtering technique is more sensitive to noise?
(A) Wiener filtering
(B) Inverse filtering
(C) Gaussian smoothing
(D) Adaptive median filtering
17. Wiener filtering is considered:
(A) A deterministic process
(B) A non-linear filter
(C) A statistical approach
(D) A sharpening filter
18. Which of the following is not needed for Wiener filtering?
(A) Power spectral density of the noise
(B) Degradation function
(C) Original image
(D) Power spectral density of the signal
19. Which frequency components are most affected in inverse filtering?
(A) Low frequency
(B) High frequency
(C) Mid frequency
(D) DC component
20. In inverse filtering, the degradation function is typically represented in:
(A) Laplace domain
(B) Time domain
(C) Frequency domain
(D) Pixel domain
21. To reduce noise amplification in inverse filtering, one can:
(A) Increase brightness
(B) Ignore low frequencies
(C) Use a threshold on the frequency response
(D) Use a larger kernel size
22. Which filter works better with images corrupted by blur and noise?
(A) Gaussian filter
(B) Inverse filter
(C) Wiener filter
(D) Sobel filter
23. Inverse filtering assumes:
(A) Random degradation
(B) Unknown blur kernel
(C) Known degradation model
(D) No blurring in the image
24. Wiener filtering can be implemented using:
(A) Convolution in spatial domain
(B) Differentiation in time domain
(C) Multiplication in frequency domain
(D) Integral transform in pixel domain
25. The mathematical foundation of Wiener filtering is based on:
(A) Bayesian estimation
(B) Neural networks
(C) Fuzzy logic
(D) Histogram equalization
26. A key limitation of inverse filtering is:
(A) Low computational cost
(B) Over-smoothing
(C) Sensitivity to noise
(D) Lack of linearity
27. Wiener filter requires:
(A) No assumptions about the noise
(B) A known original image
(C) Estimates of signal and noise power spectra
(D) Only the blurred image
28. Which technique generalizes inverse filtering to include noise consideration?
(A) Gradient filter
(B) Wiener filter
(C) Homomorphic filter
(D) Butterworth filter
29. When applied correctly, Wiener filtering:
(A) Always produces sharper images
(B) Ignores blur function
(C) Balances deblurring and noise suppression
(D) Produces binary output
30. In frequency domain, inverse filtering corresponds to:
(A) Division by the degradation function
(B) Multiplication with the noise
(C) Convolution with the kernel
(D) Subtraction of the blur
31. Wiener filtering can be interpreted as:
(A) Maximum a posteriori estimation
(B) Mean filtering
(C) Convolution with a Gaussian
(D) Histogram stretching
32. Which type of blur is often addressed using inverse and Wiener filters?
(A) Motion blur
(B) Speckle noise
(C) Quantization noise
(D) Impulse noise
33. The inverse filter is ideal when:
(A) Blur is strong and noise is high
(B) Degradation is known and noise is absent
(C) The histogram is uniform
(D) Edge detection is needed
34. Which statement is true about Wiener filtering?
(A) It ignores degradation model
(B) It requires frequency domain representation
(C) It cannot handle noise
(D) It reduces contrast in all cases
35. Which transform is commonly used before applying Wiener filtering?
(A) Laplace transform
(B) Discrete Fourier Transform
(C) Discrete Cosine Transform
(D) Wavelet Transform
36. Inverse filtering is not suitable when:
(A) Noise level is zero
(B) The degradation function is perfectly known
(C) The image has significant noise
(D) The image is grayscale
37. The Wiener filter provides better performance than inverse filtering when:
(A) Noise is absent
(B) Noise is random and measurable
(C) Image is binary
(D) Blur is due to histogram stretching
38. Wiener filtering handles the tradeoff between:
(A) Resolution and color depth
(B) Noise suppression and edge preservation
(C) Histogram matching and contrast
(D) Brightness and saturation
39. Which filter requires both noise and signal power spectrum estimation?
(A) Median filter
(B) Inverse filter
(C) Wiener filter
(D) Laplacian filter
40. The degradation function in inverse and Wiener filtering is usually:
(A) A spatial mask
(B) A histogram curve
(C) A point spread function (PSF)
(D) A threshold function
41. Which of the following filters is based on statistical theory?
(A) Gaussian filter
(B) Inverse filter
(C) Wiener filter
(D) Laplacian filter
42. The Wiener filter output improves when:
(A) Noise power is very high
(B) Signal-to-noise ratio increases
(C) Kernel size decreases
(D) Image brightness reduces
43. Wiener filtering can be applied in which domain?
(A) Time only
(B) Frequency only
(C) Spatial and frequency domains
(D) Histogram domain only
44. Inverse filtering cannot reconstruct the image well if:
(A) Noise is zero
(B) Degradation function is perfect
(C) Degradation function is close to zero in some frequencies
(D) Original image is known
45. What is typically needed before applying inverse filtering?
(A) Contrast stretching
(B) Knowledge of PSF
(C) Histogram equalization
(D) Denoising
46. The Wiener filter is adaptive in nature because it:
(A) Changes with color
(B) Adapts based on SNR
(C) Uses the Laplacian
(D) Inverts histogram
47. In real-world applications, Wiener filter is preferred over inverse filter because:
(A) It’s faster
(B) It better handles noise
(C) It needs no model
(D) It performs compression
48. Inverse filtering works on the principle of:
(A) Addition
(B) Division
(C) Subtraction
(D) Thresholding
49. Wiener filter performance improves with:
(A) More noise
(B) Less accurate PSF
(C) Better noise estimates
(D) Larger kernel
50. Which of the following is typically not a challenge for Wiener filtering?
(A) Unknown PSF
(B) Noise estimation
(C) Image scaling
(D) Power spectrum calculation
