|
The purpose of the thesis is to propose a visual recognition system combineneuron with fuzzy theory. Visual recognition techniques applied industrial productuon inspection hasbeen using for many years, but most of the inspect techniques can tell us onlythe product is good or bad, they can't recognize if the product is in betweengood and bad. In order to increase the recognition capability, we apply theNeuron and Fuzzy theory to solve this problems. Finally, we implement our thoery to robot arm, the simulation resultsverify that after this design, the recognition capability can really beincreased and improved.
|