In this chapter we've seen the power of neural networks in recognizing digits from 0 to 9 in images. Although the coding of the digits was very small in 10x10 images, it was important to understand the concept in practice. Neural networks are capable of learning from data, and provided that real-world representations can be transformed into data, it is reasonable to take into account that character recognition can be a very good example of the application in pattern recognition. The application here can be extended to any type of characters, under the condition that the neural network should all be presented with the predefined characters.