Article Directory
A profile image and histogram
1. Basic principles
Edge extraction image, the common way is to carry out gradation processing on the image, and then use the extracted frame image gradient algorithm.
Histogram is for statistical data, will organize a series of statistics to achieve defined bin. bin values are calculated from the statistic characteristics of the data, the data can be gradient direction, color, or any other feature.
2. Code
3. Results
Two Gaussian filtering
1. Basic principles
Gaussian filter is essentially a signaling, which uses a smoothing of the signals, it is known for the digital image post-application, the noise is the biggest problem, since transmission errors will be accumulated and other reasons, many image processing textbooks will very early introduction Gauss filter for obtaining a high image ratio SNR (signal reflect true). Related to this there Gauss-Laplace transformation, in fact, in order to obtain a better image edge, to make the image Gauss filtering, noise is removed, and then of Second derivative vector, determining an edge with the zero crossing of the second derivative in the calculation when the frequency domain is a product => spatial convolution. Is to establish a mathematical model filter, the image data to the energy conversion by the model, the noise is a high frequency portion is to reduce the influence of noise after smoothing Gaussian filter.
2. Code
3. Results
Three histogram equalization
1. Basic principles
Histogram equalization process "central idea" is the gray-level histogram of the original image from a gray scale more concentrated within the range becomes uniform distribution of the entire gray scale. Histogram equalization is nonlinear stretching of the image, re-assigned image pixel value, the number of pixels in a certain gradation range is substantially the same.
2. Code
3. Run results
