Machine vision systems can be used in many industrial areas. In the following overview, these areas are divided according to the nature of the problem you want to solve. Choose whether you are interested in color diagnostics, counting products, measuring their dimensions or the very widespread reading of barcodes today. For a number of areas, you will find not only examples of applications, but also a more detailed mathematical description of the image analysis function that is central to the issue.

Digital painting

The cornerstone of the whole issue of machine vision is digital image. Everyone knows him, everyone meets him every day, we take digital photos on the ground, underground, underwater, in the air, "selfie". But what is hidden behind the clutter of colors, and what is the point of distinguishing how color is formed? We will answer this in this article, which should help you see the pixels where you have seen the object so far.

A digital image is essentially a two-dimensional network of values ​​representing the light intensity of its individual points. The digital image can be expressed mathematically as a so-called image function of two variables formed by a matrix of values.

The values of f (x, y) of the picture function correspond to the brightness of the point (x, y) , where x, y represent the coordinates of the given picture element). According to established conventions, the pixel with coordinates (0,0) is located in the upper left corner of the image and the x values ​​increase from left to right and the y values ​​increase from top to bottom. The sensors used to acquire the digital image convert the monitored image into a discrete number of pixels, with each pixel being assigned a numeric location and a gray or color scale value specifying its brightness or color. The digitized image thus obtained can be characterized by three basic properties. These are the resolution, bit depth and number of color planes in the image.

The image resolution indicates the number of columns and rows of pixels in the image. An image composed of M columns (M pixels along its horizontal axis) and N rows (N pixels along its vertical axis) therefore has a resolution of MxN . Bit depth expresses the number of bits used to decode a pixel value. If the bit depth is k, then one pixel can take 2k different values. In professional applications, a color depth of 8 bits is usually used for black and white images (256 shades of gray) and 24 bits for color. Number of color planes in the image describes how many pixel networks together form a complete image. Grayscale or monochrome images consist of only one color layer, while true-color images are composed of three layers corresponding to three primary colors: red, green, and blue. Based on the bit depth and the number of color layers of an image, three basic types of images are distinguished: grayscale image, color image, and complex image. The characteristics of individual types of images, including an explanation of the system of their coding, are given in the table.

Read more:

https://promosimple.com/ps/1aeb6/machine-vision-systems-design

https://promosimple.com/ps/1aeb5/machine-vision-types-and-application

https://promosimple.com/ps/1aeb4/spring-photos-2022

https://promosimple.com/ps/1aeb3/low-perspective-photos