What are the algorithms for image processing?

According to the number of processing objects, the operation of the basic algorithm of image processing can be divided as follows:

1) point operation: the operation of processing point cell information.

2) Group operation: it deals with the operation of group units (the set of several adjacent points).

1. Binary operation

Image binarization is a very common and important operation in image processing, and it is a process of converting gray-scale images into binary images or gray-scale images. There are many binarization operations, such as general binarization, flip binarization, truncated binarization, return-to-zero binarization and return-to-zero inversion binarization.

2. Histogram processing

Histogram is another important process in image processing, which reflects the statistical information of different pixel values in the image. From this sentence, we can know that histogram information only reflects gray statistical information, and has nothing to do with the specific position of pixels. This important feature plays a key role in many recognition algorithms.

3. Template convolution operation

Template operation is often used in image processing, and many operations can be summed up as template operation, such as smoothing, filtering and edge feature extraction. What needs to be explained here is that the template used in template operation is usually a matrix of NXN (n is usually an odd number such as 3, 5, 7, ...). If this matrix is symmetric, this template is also called a convolution template, and if it is asymmetric, it is a general operation template. The templates we usually use are generally convolution templates. For example, Sobel operator template in edge extraction.