I have already tried to describe this scene with the help of clues on my own. Let’s look at this graphic scene of the universal recognition method together and use the hooks to expand this, so to speak, the angle between the 3 planes to 2d images of vertical planes, so to speak, bring them to one plane to describe in detail. The main thing here is that it is like a delivery “chute” of a special type and that there are clear lower faces between the side walls-objects and the lower plane, which is not considered.
I will add that two colors are important here, blue, which is the brightest of the colors, shines through the icicle, and yellow, as the color of the filter, which will cope with the glow of blue and that due to the fact that the scene is not made by hand accurately, there are details that are not visible.
Also, the blue object looks like a needle, but in fact no detail speaks of it as an object, it has a flash of yellow light with yellow sparks on its surface.
Also, it would be difficult to make a natural icicle in shape, there should be waves of ice on the surface.
The basis of computer recognition is a description of forces constantly pressing on each other, pulling towards each other, merging together, which must be put together in the way shown in the figure and make a detailed theoretical description, run through a neural network to use the filter obtained in the form of an image transmitting filter and an algorithm processing the image.
The presented graphical description of the image recognition problem is what solves the existing problems of image recognition inaccuracy.
So there are three issues of pattern recognition problem:
- Universal video image filter,
- Determining the luminescence, objects and dimensions of the problem,
- Scanning method: scan-build.