Answer: 1 Explanation: This is a Sobel filter. Refer to this link to know more.
Q2. Which of the following OpenCV functions can be used to threshold an image?
cv2.threshold()
cv2.thresh()
cv2.Thresh()
cv2.Threshold()
Answer: 1 Explanation: In OpenCV, cv2.threshold() can be used to threshold an image. Refer to this link to know more.
Q3. In a high contrast image, _________?
everything looks nearly of the same intensity (dull, washed-out grey look)
everything looks clear to us
Answer: 2 Explanation: In a low contrast image, it is difficult to identify details because everything looks nearly of the same intensity (dull, washed-out grey look) as compared to a high contrast image where everything looks crisp clear. Refer to this link that has examples of both low and high contrast images.
Q4. Which technique is used to construct a color image from a Bayer image?
Sampling
Interpolation
Segmentation
Quantization
Answer: 2 Explanation: Because Bayer filter uses 1 channel containing 25%-Red, 50%-Green, and 25%-Blue pixels instead of 3 channels, so to form a color image we use interpolation to find the missing information. To know more about Bayer filter, refer to this link.
Q5. Which of the following OpenCV functions can be used to create a structuring element?
cv2.getStructuringElement()
cv2.createStructEl()
cv2.createStructuringElement()
cv2.getStructEl()
Answer: 1 Explanation: In OpenCV, cv2.getStructuringElement() can be used to create a structuring element. Refer to this link to know more.
Q6. How to calculate the threshold value for adaptive Thresholding?
Mean of the pixel neighborhood
Weighted Mean of the pixel neighborhood
Median of the pixel neighborhood
Can be any of the above three
Answer: 4 Explanation: We can use any of the above mentioned statistics to calculate the threshold value for adaptive Thresholding.
Q7. In the expression g(x,y) = (k+1)*f(x,y) – k*b(x,y), if we set k=1 what is this technique called? Suppose f(x,y) is the input image, b(x,y) is its blurred version and g(x,y) is the output image.
Unsharp Masking
Highboost filtering
Laplacian of Gaussian
Difference of Gaussian
Answer: 1 Explanation: This is known as Unsharp Masking. Refer to this link to know more.
Q8. Which of the following set operations are generally used in Morphological image processing?
Union
Intersection
Difference
All of the above
Answer: 4 Explanation: All of the above set operations can be used in Morphological image processing.
Q1. Difference between the closing and the input image is known as ________?
Black top-hat transform
White top-hat transform
Hit or Miss
Closing
Answer: 1 Explanation: Difference between the closing and the input image is known as Black top-hat transform. Refer to this link to know more.
Q2. Which of the following Morphological operations output the object boundaries?
Erosion
Dilation
Morphological Gradient
Closing
Answer: 3 Explanation: As we know Morphological Gradient is the difference between the dilation and erosion of an image. Because dilation and erosion mostly affect the pixels that are close to the boundary between the foreground and background, their difference generally yields the boundary and thus this is used for edge detection and segmentation tasks.
Q3. What type of filters can be used for noise removal?
Low Pass filters
High Pass filters
Answer: 1 Explanation: Because low pass filters blocks the high-frequency parts of an image, and noise is a high frequency component so Low Pass filters can be used for noise removal.
Q4. Does using a separable filter reduces computational cost?
Yes
No
Answer: 1 Explanation: For instance, the Gaussian kernel is linearly separable. Because of this, the computational complexity is reduced from O(n^2) to O(n).
Q5. In contrast stretching, by changing the shape of the transformation function we can obtain/perform ________?
Min-Max stretching
Percentile stretching
Thresholding function
All of the above
Answer: 4 Explanation: Because Contrast Stretching uses a Piecewise Linear function so by changing the location of points we can obtain all the above mentioned functions.
Q6. Which technique expands the range of intensity levels in an image so that it spans the full intensity range of the display?
Contrast Stretching
Gamma Correction
Intensity Level Slicing
None of the above
Answer: 1 Explanation: Out of the above mentioned techniques, Contrast Stretching expands the range of intensity levels in an image.
Q7. Which of the following is an example of Affine Transformation?
Translation
Rotation
Scaling
All of the above
Answer: 4 Explanation: All of the above are examples of Affine Transformation as they preserves collinearity, parallelism as well as the ratio of distances between the points.
Q8. Which of the following is an additive color model?
RGB
CMYK
Both of the above
None of the above
Answer: 1 Explanation: In additive model the colors present in the light add to form new colors. For instance, in RGB color model, Red, Green, and Blue are added together in varying proportions to produce an extensive range of colors. To know more about additive models, refer to this link.
Q1. A color image is formed by a combination of ________ colors?
Red, Green, and Blue
Black, and White
Red, Green, and White
All colors
Answer: 1 Explanation: A color image has 3 channels that contains the intensity information corresponding to three wavelengths red, green, and blue (RGB) collectively called primary colors or channels.
Q2. In spatial filtering, we convolve the neighborhood of a pixel with a subimage. What we generally call this subimage?
mask
kernel
filter
All of the above
Answer: 4 Explanation: We refer to this subimage by all of the above mentioned names.
Q3. Which of the following interpolation algorithms results in a blocky or pixelated image?
Nearest Neighbor
Bilinear
Bicubic
None of the above
Answer: 1 Explanation: Because in Nearest Neighbor we are only replicating the nearest pixel value so this results in a blocky or pixelated image. To know more about Nearest Neighbor, refer to this link.
Q4. Does Sobel filter have some smoothing effect?
Yes
No
Answer: 1 Explanation: Because Sobel filter is obtained by multiplying the x, and y-derivative filters with some smoothing filter(1D) so this has some smoothing effect.
Q5. Which of the following arithmetic operations can be used for Image Enhancement?
Image Averaging
Image Subtraction
Image Multiplication
All of the above
Answer: 4 Explanation: All the arithmetic operations such as Averaging, subtraction or multiplication can be used for image enhancement. For instance, Averaging can be used for noise removal, subtraction can be used for detecting blockage or moving objects and multiplication can be used for extracting region of interest (roi) using masking etc.
Q6. Suppose we plot the 2D color histogram for Red and Green channels. So, what does each point within this histogram represents?
Frequency corresponding to each Red and Green pair
Pixel location corresponding to each Red and Green pair
Intensity value of the Red color
Intensity value of the Green color
Answer: 1 Explanation: For the above case, Y and X-axis correspond to the Red and Green channel ranges( for 8-bit, [0,255]) and each point within the histogram shows the frequency corresponding to each Red and Green pair. To know more about 2D Image Histograms, refer to this link.
Q7. A 0 degree hue results in which color?
Red
Green
Blue
White
Answer: 1 Explanation: A 0 degree hue results in Red color. Refer to this link to know more.
Q8. What response does a first order derivative filter will give for flat regions in an image?
Zero
positive
negative
Answer: 1 Explanation: As we know that the differentiation of a constant is 0. Since the flat region has nearly constant intensity so the first order derivative filter will give 0 response.
a filter which can be written as a product of two more simple filters
a filter that can separate noise from other features
a filter which can be written as a sum of two more simple filters
There is no such term!!!
Answer: 1 Explanation: A separable filter is the one that can be written as a product of two more simple filters.
Q2. Which of the following OpenCV functions can be used to perform Adaptive Thresholding?
cv2.adaptiveThreshold()
cv2.threshold()
cv2.adaptThreshold()
cv2.thresh()
Answer: 1 Explanation: In OpenCV, cv2.adaptiveThreshold() can be used to perform Adaptive Thresholding. Refer to this link to know more.
Q3. In some cases, the Adaptive Histogram Equalization technique tends to over-amplify the noise. Which technique is used to solve this problem?
Histogram Specification
CLAHE
SWAHE
All of the above
Answer: 2 Explanation: The Adaptive Histogram Equalization (AHE) technique tends to over-amplify the noise so to avoid this contrast limiting is applied and this method is known as Contrast Limited Adaptive Histogram Equalization (CLAHE).
Q4. What is Ringing effect in image processing?
a rippling artifact near sharp edges
a rippling artifact in smooth areas
In this, the filter rings(warns) about the noise
There is no such effect!!!
Answer: 1 Explanation: In image processing, ringing effect refers to a rippling artifact near sharp edges. To know more about this effect, refer to this link.
Q5. What is local contrast enhancement?
In this, we divide the image into small regions and then perform contrast enhancement on these regions independently
In this, we divide the image into small regions and then perform contrast enhancement on all these regions using same transformation function
In this, we simply perform contrast enhancement on the entire image
None of the above
Answer: 2 Explanation: As clear from the name, in local contrast enhancement we divide the image into small regions and then perform contrast enhancement on these regions independently. The transformation function for this is derived from the neighborhood of every pixel in the image.
Q6. In the Gaussian filter, what is the relation between standard deviation and blurring?
Larger the standard deviation more will be the blurring
Larger the standard deviation less will be the blurring
No relation!!!
Answer: 1 Explanation: In Gaussian filter, larger the standard deviation more will be the blurring.
Q7. Which of the following are the common uses of image gradients?
Edge detection
Feature Matching
Both of the above
None of the above
Answer: 3 Explanation: Image gradients can be used for both Edge Detection (for instance in Canny Edge Detector) and Feature Matching.
Q8. How does the change in filter size affects blurring? Assume the filter is a smoothing filter.
Blurring increases with decrease in the filter size
Blurring decreases with increase in the filter size
Blurring increases with increase in the filter size
There is no effect of filter size on blurring!!!
Answer: 3 Explanation: As we increase the filter size, Blurring also increases.
Answer: 1 Explanation: In Lab, L* stands for perceptual lightness, and a* and b* for the four unique colors of human vision: red, green, blue, and yellow. Refer to this link to know more.
Q2. In Nearest Neighbor, how many neighboring pixels are considered for calculating the intensity value for a new location?
3
1
2
0
Answer: 2 Explanation: As clear from the name, this method considers the nearest neighbor i.e. 1 pixel for calculating the intensity value for a new location. To know more about Nearest Neighbor, refer to this link.
Q3. _______ is used to keep the output image size similar to the input image during convolution operation?
Padding
Interpolation
Dilation
Erosion
Answer: 1 Explanation: Padding refers to the process of adding borders in an image with generally 0 valued pixels. Because during convolution the size of the image decreases so to prevent this we pad the original image.
Q4. What do you mean by Affine Transformation?
a geometric transformation that preserves collinearity and parallelism
a geometric transformation that preserves distances and angles but not collinearity
transformation that is associated with the change in viewpoint
a geometric transformation that preserves collinearity and distance but not parallelism
Answer: 1 Explanation: An affine transformation is any transformation that preserves collinearity, parallelism as well as the ratio of distances between the points (e.g. midpoint of a line remains the midpoint after transformation). It doesn’t necessarily preserve distances and angles. Refer to this link to know more.
Q5. Can we sharpen an image using a smoothing filter?
Yes
No
Answer: 1 Explanation: Yes, we can sharpen an image using a smoothing filter. For instance, both Unsharp Masking and Difference of Gaussian techniques both sharpen an image using a smoothing filter.
Q6. What do you mean by Domain filters?
in which the filter weights are assigned according to the spatial closeness
in which the filter weights are assigned according to the intensity difference
in which the filter weights are assigned both according to the spatial closeness and intensity difference
Answer: 1 Explanation: As clear from the name, Domain filters are the one in which the filter weights are assigned according to the spatial closeness (i.e. Domain)
Q7. Which of the following histogram techniques can be used for image segmentation?
Histogram Equalization
CLAHE
Histogram Backprojection
Histogram Specification
Answer: 3 Explanation: Histogram Backprojection can be used for image segmentation. To know more about Histogram Backprojection, refer to this link.
Q8. In general, the gradient in x-direction will find ________?
Horizontal edges
Vertical edges
Any type of edges
Gradient has no relation with edges
Answer: 2 Explanation: Because gradient refers to the directional change in intensity so the gradient in x-direction will find Vertical edges. Refer to this link to know more.
Q1. Which of the following denotes 255 in the binary form?
11111111
00000000
01111111
11111110
Answer: 1 Explanation: In binary form, 255 is represented as 11111111. To know more about how to convert decimal to binary, refer to this link.
Q2. Which of the following OpenCV functions can be used to perform CLAHE?
First we create a CLAHE object using “cv2.createCLAHE()” and then we apply this on the image using .apply() method
cv2.applyCLAHE()
cv2.clahe()
None of the above
Answer: 1 Explanation: In OpenCV, first we create a CLAHE object using “cv2.createCLAHE()” and then we apply this on the image using .apply() method. Refer to this link to know more about this function.
Q3. What is the smallest element of an image?
pixel
dpi
meter
centimeter
Answer: 1 Explanation: In digital image processing, a pixel(or picture element) is the smallest item of information in an image.
Q4. Which of the following OpenCV functions can be used to apply an affine transformation to an image?
cv2.warpAffine()
cv2.affineTransform()
cv2.applyAffine()
cv2.WarpAffine()
Answer: 1 Explanation: In OpenCV, cv2.warpAffine() can be used to apply an affine transformation to an image. Refer to this link to know more.
Q5. Which of the following is an subtractive color model?
RGB
CMYK
Both of the above
None of the above
Answer: 2 Explanation: In subtractive model colors are perceived as a result of reflected light. For instance, Cyan is the complement of red, meaning that the cyan serves as a filter that absorbs red. The amount of cyan applied to a white sheet of paper controls how much of the red in white light will be reflected back from the paper. To know more about subtractive models, refer to this link.
Q6. What type of filters results in image sharpening?
Low Pass filters
High Pass filters
Answer: 2 Explanation: Because high pass filters enhances the high-frequency parts of an image, these results in image sharpening.
Q7. For a skewed image histogram, which technique will can be used for improving the global contrast?
Histogram Equalization
Histogram Matching
Histogram Balancing
None of the above
Answer: 2 Explanation: For skewed image histogram, one reasonable approach can be to manually specify the transformation function that preserves the general shape of the original histogram but has a smoother transition of intensity levels in the skewed areas. This is what we do in Histogram Matching.
Q8. What does the term “Shadows” refers to in a 1D image histogram?
Leftmost part (the black and dark areas)
Rightmost part (light and pure white areas)
Center part (medium grey areas)
There is no such term!!!
Answer: 1 Explanation: Shadows as clear from the name refers to the Leftmost part of the histogram that contains mostly the black and dark areas. To know more about Image Histograms, refer to this link.
Answer: 1 Explanation: The main steps used in Canny Edge Detector are Noise Reduction, Finding Intensity Gradient, Non-max Suppression, Hysteresis Thresholding. Refer to this link to know more.
Q2. Which of the following filter assigns more weights to the nearest pixels as compared to the distant pixels?
Gaussian Filter
Box Filter
Median Filter
All of the above
Answer: 1 Explanation: A Gaussian filter assigns more weights to the nearest pixels as compared to the distant pixels. To know more about Gaussian filter, refer to this link.
Q3. How many thresholds are used in hysteresis thresholding in Canny Edge Detector?
2
1
3
4
Answer: 1 Explanation: To solve the problem of “which edges are really edges and which are not” Canny uses the Hysteresis thresholding. In this, we set two thresholds ‘High’ and ‘Low’. Refer to this link to know more.
Q4. In Bilinear Interpolation, how many neighboring pixels are considered for calculating the intensity value for a new location?
3
1
2
4
Answer: 4 Explanation: Bi-linear interpolation means applying a linear interpolation in two directions. Thus, it uses 4 nearest neighbors for calculating the intensity value for a new location. To know more about Bilinear Interpolation, refer to this link.
Q5. Which of the following image sharpening techniques to use if the image contains a high degree of noise?
Sobel
Laplacian
Difference of Gaussian
Scharr
Answer: 3 Explanation: Because in Difference of Gaussian we are actually doing blurring which reduces the effect of noise to a great extent.
Q6. What response does a second order derivative filter will give along the ramps?
Zero
positive
negative
Answer: 1 Explanation: Because ramps has constant slope so for ramp edge the first order derivates gives constant response along the ramp while second order gives 0 response along the ramp.
Q7. Which of the following OpenCV functions can be used to perform Hit-or-Miss Transform?
Answer: 1 Explanation: In OpenCV, cv2.morphologyEx(img, cv2.MORPH_HITMISS, kernel) can be used to perform Hit-or-Miss Transform. Refer to this link to know more.
Q8. Which of the following Morphological operations closes the holes/gaps present in the object while keeping the initial object size same?
Dilation
Erosion
Closing
Opening
Answer: 3 Explanation: As clear from the name, Dilation dilates or expands the object region so for bridging/closing the holes Dilation can be used but to keep the initial object size same we need Erosion. Since Closing is Dilation followed by Erosion so this can also be used.
Q1. In an 8-bit color image, the intensity value (255,255,255) corresponds to which color? (Consider the RGB color model here)
Black
White
Red
Cyan
Answer: 2 Explanation: Because RGB is an additive color model i.e. the colors present in the light add to form new colors. So, Zero intensity for each component(RGB) gives the darkest color (no light, considered the black), and full intensity of each gives a white. Since for an 8-bit image the full intensity value is 255 so the answer is White. Refer to this link to know more.
Q2. Which of the following techniques can be used for blur detection or detecting blurred images?
Variance of Laplacian
Unsharp Masking
High Boost filtering
All of the above
Answer: 1 Explanation: As we all know that the blurry image doesn’t have well-defined edges. So, if you calculate the Laplacian of this image, you will get more or less the same response everywhere. In other words, the variance of this Laplacian image will be less. So, this can be used for blur detection.
Q3. Which of the following can be a reason for a low contrast image?
Poor illumination of the scene
wrong setting of lens aperture during image acquisition
lack of dynamic range in the imaging sensor
All of the above
Answer: 4 Explanation: All of the above can be possible reasons for getting a low contrast image.
Q4. For which type of images, the Histogram Equalization technique can be used?
thermal images
Satellite images
X-ray images
All of the above
Answer: 4 Explanation: Because Histogram Equalization is a contrast enhancement method so this can be used in all of the above images.
Q5. Which of the following OpenCV functions can be used to perform the Dilation operation?
cv2.dilate()
cv2.Dilate()
cv2.dilate2D()
cv2.morphDilate()
Answer: 1 Explanation: In OpenCV, cv2.dilate() can be used to perform Dilation operation. Refer to this link to know more.
Q6. Generally in a 1D image histogram, what do the x and y-axis represents?
x-axis: Intensity values, Y-axis: no. of pixels corresponding to intensity values
x-axis: no. of pixels corresponding to intensity values, Y-axis: Intensity values
x-axis: pixel location, y-axis: Intensity values
x-axis: Intensity values, y-axis: pixel location
Answer: 1 Explanation: In 1D image histogram, we plot the intensity values on the x-axis and the no. of pixels corresponding to intensity values on the y-axis. To know more about Image Histograms, refer to this link.
Q7. What of the following is the general form of representation of log transformation? Suppose r and s denote the input and output pixel values respectively.
s=clog10(1/r)
s=clog10(1+r)
s=clog10(1-r)
s=clog10(1*r)
Answer: 2 Explanation: Log transformation means replacing each pixel value with its logarithm. The general form of log transformation function is given by s=clog10(1+r). Refer to this link to know more.
Q8. What is a high pass filter?
a filter that enhances the high-frequency parts of an image
a filter that enhances the low-frequency parts of an image
Answer: 1 Explanation: A high pass filter is the one that enhances the high-frequency parts of an image.
Q1. Which of the following Morphological operations closes the holes/gaps present in the object while keeping the initial object size the same?
Dilation
Erosion
Closing
Opening
Answer: 3 Explanation: As clear from the name, Dilation dilates or expands the object region so for bridging/closing the holes Dilation can be used but to keep the initial object size same we need Erosion. Since Closing is Dilation followed by Erosion so this can also be used.
Q2. Which of the following OpenCV functions can be used to threshold an image?
cv2.threshold()
cv2.thresh()
cv2.Thresh()
cv2.Threshold()
Answer: 1 Explanation: In OpenCV, cv2.threshold() can be used to threshold an image. Refer to this link to know more.
Q3. Difference between the image and its opening is known as __________?
Black top-hat transform
White top-hat transform
Hit or Miss
Closing
Answer: 2 Explanation: Difference between the image and its opening is known as White top-hat transform. Refer to this link to know more.
Q4. What is the contrast in image processing?
the difference in luminance or color that makes an object distinguishable from other objects within the same field of view.
the difference in the resolution that makes an object distinguishable from other objects within the same field of view.
same as brightness
There is no such term in image processing!!!
Answer: 1 Explanation: In image processing, contrast refers to the difference in luminance or color that makes an object distinguishable from other objects within the same field of view. To know more about contrast, refer to this link.
Q5. Which of the following refers to a type of noise in an image?
Gaussian
Salt and Pepper
Speckle
All of the above
Answer: 4 Explanation: All of the above refers to a type of noise in an image.
Q6. Which of the following Morphological operations can be used for shape detection or finding particular patterns in the given image?
Morphological Gradient
Hit-or-Miss Transform
Top-hat Transform
Opening
Answer: 2 Explanation: Hit-or-Miss Transform can be used for shape detection or finding particular patterns in the given image. In this, we use two structuring elements (say B1 and B2) and ask a simple question of does B1 fits the object while, simultaneously, B2 misses the object, i.e. fits the background. Refer to this link to know more.
Q7. Which of the following matplotlib functions can be used to calculate the image histogram? Suppose we import as “import matplotlib.pyplot as plt”
plt.hist()
plt.calcHist()
plt.showHist()
plt.histCalc()
Answer: 1 Explanation: In matplotlib, plt.hist() can be used to calculate the image histogram. Refer to this link to know more about this function.
Q8. In Perspective Transformation, what is the minimum number of points to select to obtain the transformation matrix?
1
2
3
4
Answer: 4 Explanation: In Perspective Transformation, the transformation matrix (M) is defined by 8 constants, thus to find this matrix we need 4 points. Refer to this link to know more.
Q1. Which domain refers to the Fourier transform of an image?
Spatial domain
Frequency domain
Answer: 2 Explanation: In transform domain we first transform an image into another domain (like frequency) by applying for instance Fourier transform, do processing there and convert it back to the spatial domain by some inverse operations.
Q2. Which of the following techniques can be used for image segmentation?
Histogram Equalization
CLAHE
Histogram Backprojection
Histogram Specification
Answer: 3 Explanation: Histogram Backprojection can be used for image segmentation. To know more about Histogram Backprojection, refer to this link.
Q3. Which of the following OpenCV functions can be used to perform convolution operations?
cv2.filter2D()
cv2.convolve()
cv2.filter()
cv2.conv2D()
Answer: 1 Explanation: In OpenCV, cv2.filter2D() can be used to perform convolution operation. Refer to this link to know more about this function.
Q4. Which color model is used in printing?
Additive color model
Subtractive color model
Answer: 2 Explanation: Subtractive color model (CMYK) is used in printing. Refer to this link to know more.
Q5. Which domain refers to the image plane itself?
Spatial domain
Frequency domain
Answer: 1 Explanation: Spatial domain refers to the image plane. This means we perform all operations directly on image pixels.
Q6. In the dilation operation, generally the output image features becomes ________ ?
Thinner
Thicker
Blurred
Sharpened
Answer: 2 Explanation: Because Dilation dilates or expands the object region, the output image features becomes thicker. Refer to this link to know more.
Q7. Dilation is the ________ of Erosion?
dual
rotated version
translated version
neighbor
Answer: 1 Explanation: Dilation is the dual of erosion. Dual in the sense that dilating the object region is equivalent to eroding the background region and vice versa.
Q8. What type of filters results in image sharpening?
Low Pass filters
High Pass filters
Answer: 2 Explanation: Because high pass filters enhances the high-frequency parts of an image, these results in image sharpening.