Zero Mind Map Analysis
Use the mind map to briefly preview the reading notes taken when reading the visual part.
1 Basic concepts
1.1 Image processing and image analysis
Before image analysis, the image needs to be pre-processed. After all, some pictures have a greater impact of noise and interference.
- Image processing: A general term for methods and techniques that improve, simplify, enhance, and change images.
- Image analysis: The process of analyzing captured and processed images to extract image information, identify objects, or extract features about objects in the image or the surrounding environment.
1.2 Two-dimensional and three-dimensional images
- Applications of 2D images: feature extraction, detection, navigation, part processing and many others.
- Applications of three-dimensional images: applications in motion detection, depth measurement, remote sensing, relative positioning and navigation.
The book also mentions that all three-dimensional vision systems have the same problem-how to handle the many-to-one mapping from scene to image.
1.3 The nature of images
The book uses a printer as an example to adjust the color presentation according to the proportion of black and white ink sprayed.
Electronic and digital images are often divided into many small pieces, each called a primitive or pixel (or volume element or voxel in three-dimensional images), all of which are the same size.
An image file is simply a large collection of pixel intensity data. The essence of the color image is the same, the difference is that the original color image is divided into three images: red, green and blue before being acquired and digitized.
1.4 Acquisition of images
There are two common types of cameras: analog and digital cameras. Analog cameras are not commonly used, so the focus is on digital cameras.
The acquired images will be digitized and turned into a collection of 0s and 1s.
1.5 Digital images
A digital image contains intensity values (sets of 0s and 1s) stored sequentially at individual pixels.
If the system is a four-digit digital system, one pixel can have 2 to the 4th power, that is, 16 different intensities.
The program will read the intensity value of each pixel in four-bit units.
Grayscale image: Each pixel in an image has a different grayscale level
Color image: 3 images with red, green and blue (RGB) tones are combined to obtain
Binary image: either 0 or 1, that is, either white or black
In order to obtain a binary image, in most cases the histogram and threshold of the grayscale image are used.
1.6 Frequency domain and spatial domain
Many image processing and analyzes are based on the frequency domain or spatial domain.
Frequency domain processing: Use the frequency of the image to change, analyze and process the image, and process the frequency representation of the entire image.
Spatial domain processing: The operation object is a single object, and the operation process directly affects each pixel.