SIMPLE REGIONAL DESCRIPTORS

Hi, this is all lucky Devine, today I'm going to talk about

simple

regional

descriptors

. What are

regional

descriptors

? Regional descriptors are used to describe image regions. There are several approaches to describing image regions. These are

simple

descriptors. Topological descriptors. Texture and momentum invariance. descriptors Simple descriptors are often called basic descriptors. These are parameters of area, circularity ratio, compactness range, mean and median. Other statistical descriptors, such as variance, standard deviation, and coefficient of variation, are also called simple descriptors, although these descriptors look simple but go a long way. of information about the region of interest in images the number of pixels in the relationship of interest is known as the area the length of the boundary of the region in the image is the perimeter although area and perimeter are not always used because they will change when the region the scale is variable, these two are mainly used in the calculation of other receivers, such as the circularity ratio and the compactness area and the perimeter are not scale variants, the compactness and the circularity ratio the compactness of a region is defined as the perimeter Square by area the measure of compactness of a shape is a numerical quantity that represents the degree to which the shape is compact circularity relation is the relationship between the region and the area of ​​the circle that has the same perimeter as that of the region of interest circularity ratio RC is given by 4 Pi a times P squared here a is the area and P is the compactness parameter and the circularity ratio of scalar multiples to each other.

Compactness and circularity ratio are dimensionless quantities, so they do not depend on the scale factor, since the area at the perimeter is calculated from the digital approximation of images may contain errors when calculating these quantities, while this Compactness and circularity ratio is calculated from the area and perimeter, so the compactness and circularity ratio can also contain the maximum and minimum intensity errors and the median maximum intensity is the maximum intensity values ​​of all the values ​​of intensity of pixels present in the region of interest similarly minimum Industries minimum of all pixel intensity values ​​in the range of the image region is the difference between the maximum and minimum intensity values ​​median is the median of all pixel intensity values ​​within the region mean and mean variance is the average of the pixel intensity values ​​within the region variance is the second moment above the mean we already know the statistical definitions of the mean and the variance the given image is the satellite image of America the brightest spots help us calculate the amount of electricity used approximately the given table is the relationship between the areas of the bright spots and the total area here in this Matlab program the bright circles and dark are the input image initially we will apply the threshold operation on the image with the threshold value as 10. region props is a function to discover the properties of the region the parameters of the function are the properties that will be calculated here the image is provided as input to the function after running the program we can see the three circular regions in the input image in the command window we can see the properties of three regions such as circularity of area and perimeter, and centroid are Calculate as said above, we can find the properties of the image regions using region probe function in Matlab, we can find the circularity ratio centroid of the area etc., we can find the maximum and minimum values ​​of the industry only for grayscale images and not for binary images, the circularity ratio is close to 1 due to the circular regions in the input image, it's all about sample descriptors, thanks