When you first started learning digital image processing, did you also have the following problems:
The opening chapter of the textbook introduces the introduction and mathematical tools, which seems to be understandable but does not seem to involve programming...
Next, we will learn grayscale transformation, spatial filtering and frequency domain filtering, which involve rich content and various methods. I tried to write a few programs but couldn't continue...
I started learning OpenCV, found a few reference books, and successfully ran the first routine. However, when faced with the explanation of dozens of algorithm functions in each module, I felt "overwhelmed" again...
Even though these algorithm functions have both explanations and routines, they are still confusing and cannot be truly understood and mastered. Therefore, learning OpenCV also faces the risk of giving up halfway...
If this is the case for you, don’t worry, it’s normal and most people will have this learning experience.
In my school, the textbook chosen in the professional innovative education course is "Digital Image Processing (Fourth Edition)" by Gonzalez.
At the beginning of my study, I encountered all the above problems and felt that the getting started process was very difficult!
Faced with this situation, my approach is: go back and honestly follow the knowledge system of digital image processing and use OpenCV to program and implement the problems in the course one by one. To this end, I also opened a blog column "youcans' OpenCV routines" on the CSDN website, and organized my notes on learning OpenCV digital image processing into a blog. So far, I have published more than 260 articles.
Through study, I deeply realized:
Learning digital image processing must be inseparable from programming practice, otherwise it won’t even be considered as a skill on paper;
To learn OpenCV, you must understand the knowledge system of digital image processing. Otherwise, you just know what it is and don’t know why. You will be very happy at first, but the more you go on, the more confused you will be.
In my opinion: Digital image processing and OpenCV are like left and right hands, they complement each other and are mutually reinforcing.
However, when I checked more than a dozen digital image processing textbooks and books about OpenCV, I found that they seemed to belong to two schools, each with its own system and distinct distinctions. Few books can integrate them into one, and it is difficult to make people like A "rookie" like me can easily get started and advance quickly.
So, I decided to write a book myself as a bridge for learning digital image processing and OpenCV!
The positioning I designed for this book is:
It is suitable for students studying digital image processing courses. It can not only be used as a course reference, but also allows you to easily get started with OpenCV.
It is suitable for beginners of OpenCV. It is easier than the routines in the official documentation and will give beginners a deeper understanding of digital image processing.
It is also suitable for readers with a certain foundation, so that they can use this book as a manual of commonly used OpenCV functions and refer to it when needed.
This is the origin of the book "Digital Image Processing: Based on OpenCV-Python"!
The main content of this book
(Scroll down to view)
This book is based on OpenCV-Python and introduces basic methods and advanced applications of digital image processing. The book has 18 chapters in total, divided into four parts.
The first part introduces the basic operations of OpenCV-Python, including Chapters 1 to 4.
◎ Chapter 1 introduces the basic operations of images, including image reading, saving and display methods.
◎ Chapter 2 introduces the data format of images, including basic methods of image creation, copying, cropping, splicing, splitting, and merging.
◎ Chapter 3 introduces color image processing, including image color space conversion, image pseudo-color processing, etc.
◎ Chapter 4 introduces drawing and mouse interaction, mainly introducing its operation methods.
The second part introduces the basic methods of image processing, including Chapters 5 to 9.
◎ Chapter 5 introduces arithmetic operations on images, including addition, subtraction, multiplication, division and bitwise operations.
◎ Chapter 6 introduces the geometric transformation of images, including image translation, image scaling, image rotation, etc.
◎ Chapter 7 introduces the grayscale transformation of images, including linear grayscale transformation and nonlinear grayscale transformation.
◎ Chapter 8 introduces image histogram processing and improves image quality by adjusting the histogram.
◎ Chapter 9 introduces image threshold processing, including OTSU threshold algorithm, multi-threshold processing algorithm, etc.
The third part introduces advanced methods of image processing, including Chapters 10 to 13.
◎ Chapter 10 introduces image convolution and spatial filtering to achieve image blur and image sharpening.
◎ Chapter 11 introduces Fourier transform and frequency domain filtering to design more abundant filters.
◎ Chapter 12 introduces morphological image processing, including basic operations, commonly used algorithms and typical applications.
◎ Chapter 13 introduces image transformation, reconstruction and restoration, including Hough transform, Leyden transform, image reconstruction and degraded image restoration, etc.
The fourth part introduces computer vision, including Chapters 14 to 18.
◎ Chapter 14 introduces edge detection and image contours, including gradient operators, LoG operators, DoG operators, etc. for edge detection.
◎ Chapter 15 introduces image segmentation, including watershed algorithm, graph cut segmentation algorithm, mean shift algorithm, etc.
◎ Chapter 16 introduces feature description, including Fourier descriptor of feature description, regional feature description of feature description, etc.
◎ Chapter 17 introduces feature detection and matching, including Harris, SIFT, SURF, FAST, ORB and MSER algorithms, etc.
◎ Chapter 18 introduces machine learning, including principal component analysis, k-means clustering algorithm, k-nearest neighbor algorithm, Bayesian classifier, support vector machine and artificial neural network algorithm, etc.
Features of this book
- Rich examples and detailed comments
This book introduces OpenCV routines and compiles a routine index, comprehensively covering the basic functions of OpenCV, and systematically introducing the content of digital image processing courses. The implementation methods of all routines are simple and clear, and the comments are detailed, making it easy for readers to understand and modify.
- Step by step and well organized
Different from similar books, this book takes the digital image processing knowledge system as the main line rather than organizing it according to the OpenCV module, so as to better reflect the problem orientation and demand orientation.
The content of each chapter is relatively independent. Each routine is an independent program and will not call each other, thus avoiding the use of functions or content that have not been mentioned yet.
- Can be used as a function manual, ready to learn and use
This book introduces commonly used OpenCV functions and compiles function indexes. Different from official documents and similar books, this book conducts a large number of tests on OpenCV functions, focuses on explaining the format requirements and precautions for parameters in the function, and combines routines to help readers understand the special regulations of each function.
- Rich download resources
The book contains 168 color illustrations, all of which are available for download, making learning more intuitive.
There are a total of 173 routine resources in the book, as well as the pictures and video resources required to run the routines. They can also be downloaded for easy operation while reading.
Finally, a little reading suggestion for readers:
This book can be read in conjunction with Gonzalez's "Digital Image Processing" textbook. Combined with the author's experience, this learning effect will be better and get twice the result with half the effort.
