Image Processing 100 Questions

---

This is the English version of image processing 100 questions. The original Japanese repository was created by yoyoyo-yo. It’s updated by him now. To be honest, I can not speak Japanese. Since the code is language independent and I’m preparing for my interview questions about computer vision now. I decided to translate the English version of this 100 questions.

Hope this could help more people and thanks for yoyoyo-yo's great effort.

I'm currently working on ROS and OpenCV, so I'm updating another repository: OpenCV_for_Beginners. Please feel free to fork and pull request.

Before Reading

1. I’ll use the Google translator to help me understand his original meaning. Pretty cool, right? It’s the time for NLP.

2. I may add some additional materials and my own opinions to this repository.

3. I’ll discard the irrelevant parts and mainly focus on the OpenCV parts.

4. I’ll update this as soon as possible since I have to do my research project now.

Environment Setting

1. Go to Miniconda website, download and install it.

2. Open your terminal, create a virtual environment using following command:

   $ conda create python = 3.6 -n Image_Processing_100

3. Activate your virtual environment:

   $ source activate Image_Processing_100

4. Install the packages:

   $ pip install -r requirement.txt

Testing Your Environment

1. Clone this repository into your local computer:

   $ git clone git@github.com:KuKuXia/Gasyori100knock.git

2. In the Image_Processing_100_Questions folder, make a new file named sample.py, copy and paste the following code:

   import cv2
   
   img = cv2.imread("assets/imori.jpg")
   cv2.imshow("imori", img)
   cv2.waitKey(0)
   cv2.destroyAllWindows()

3. Save the file and run it.

   python ./sample.py

4. Successful if the following image is displayed with a new window! Then if you press any button, it disappears.

4. Next, see the Tutorial folder to learn numpy related to image processing. (Please skip this if you already know.)

5. From now on, please solve the Question. The Question contents are included in each folder. Please use assets/imori.jpg in the question. Questions are written in the README.md of each folder.

Content

​ Question >> Folder

• Question 1 - 10 >> Question_01_10
• Question 11 - 20 >> Question_ 11 _ 20
• Question 21 - 30 >> Question_ 21 _ 30
• Question 31 - 40 >> Question_ 31 _ 40
• Question 41 - 50 >> Question_ 41 _ 50
• Question 51 - 60 >> Question_ 51 _ 60
• Question 61 - 70 >> Question_ 61 _ 70
• Question 71 - 80 >> Quesiton_ 71 _ 80
• Question 81 - 90 >> Question_ 81 _ 90
• Question 91 - 100 >> Question_ 91 _ 100

Note

• This paper is a teaching material to learn fundamental knowledge and theory of image processing.
• In the solution, we do not use main () etc. to simplify the code as much as possible.
• We will use numpy, but we do not post basic knowledge about numpy. Please check each one.

Question

Unresolved issues are unanswered

Question 1-10

Number	Question	Number	Question
1	Channel swapping	6	Discretization of Color
2	Grayscale	7	Average Pooling
3	Binarization	8	Max Pooling
4	Binarization of Otsu	9	Gaussian Filter
5	HSV Conversion	10	Median Filter

Question 11 - 20

Number	Question	Number	Question
11	Smoothing Filter	16	Prewitt Filter
12	Motion Filter	17	Laplacian Filter
13	MAX_MIN Filter	18	Emboss Filter
14	Differential Filter	19	LoG Filter
15	Sobel Filter	20	Histogram Display

Question 21 - 30

Number	Question	Number	Question
21	Histogram Normalization	26	Bi-linear Interpolation
22	Histogram Operation	27	Bi-cubic Interpolation
23	Histogram Equalization	28	Affine transformation (Translation)
34	Gamma Correction	29	Affine transformation (Scaling)
25	Nearest Neighbor Interpolation	30	Affine transformation (Rotation)v_1, v_2

Question 31 - 40

Number	Question	Number	Question
31	Affine Transformation(Skew)	36	JPEG Compression (Step 1) Discrete Cosine Transform
32	Fourier Transfrom	37	PSNR
33	Fourier Transform Low Pass Filter	38	JPEG Compression (Step 2) DCT + Quantization
34	Fourier Transform High Pass Filter	39	JPEG Compression (Step 3) YCbCr Color System
35	Fourier Transform Band Pass Filter	40	JPEG Compression (Step 4) YCbCr + DCT + Quantization

Question 41 - 50

Number	Question	Number	Question
41	Canny Edge Detection (Step 1) Edge Magnitude	46	Hough Transform / Line Detection (Step. 3) Hough Inverse Transform
42	Canny Edge Detection (Step 2) Thinning	47	Morphological Processing (Expansion)
43	Canny Edge Detection (Step 3) Hysteresis Threshold Processing	48	Morphology Processing (Erosion)
44	Hough Transform / Line Detection (Step. 1) Hough Transform	49	Opening Processing
45	Hough Transformation / Line Detection (Step. 2) NMS	50	Closing Processing

Question 51 - 60

Number	Question	Number	Question
51	Morphology gradient	56	Template matching NCC
52	Top hat conversion	57	Template matching ZNCC
53	Black hat conversion	58	Labeling 4 Near
54	Template matching SSD	59	Labeling 8 neighborhood
55	Template matching SAD	60	Alpha Blend

Question 61 - 70

Number	Problem	Number	Problem
61	4-connected number	66	HOG (Step 1) Gradient intensity · Gradient angle
62	8 - number of connections	67	HOG (Step 2) gradient histogram
63	Thinning	68	HOG (Step 3) Histogram normalization
64 not yet	Hiruditchi thinning	69	HOG (Step 4) Drawing feature quantity
65	Zhang-Suen thinning	70	Color tracking

Problem 71 - 80

Number	Problem	Number	Problem
61	4-connected number	66	HOG (Step 1) Gradient intensity · Gradient angle
62	8 - number of connections	67	HOG (Step 2) gradient histogram
63	Thinning	68	HOG (Step 3) Histogram normalization
64 not yet	Hiruditchi thinning	69	HOG (Step 4) Drawing feature quantity
65	Zhang-Suen thinning	70	Color tracking

TODO

Hough, Gabor, adaptivebinalizatino