If the images being entered are always in same condition, there’s no problem with the normal machine vision.

​

But even with the subtle changes in lighting, the detection rate is bound to vary greatly.

​

In such circumstances, there arises a need for people to put in and readjust the parameter.

​

However DL is designed to recognize and classify the object even in the midst of severe distortion of the image.

​

Since it recognizes the target at high speed, it’s also available in real-world applications.