Technical instructions.

Create a folder, named PRV, where you want, i.e. on desktop or in the folder  "program" of disc C:  Extract in it "Eseguibili e immagini".
In this folder you can find the running programs and the data bank of images. At the moment  the images  are capital letters of the Latin alphabet, named by their name: A, B, C, D, E, ........ Z and have the .PIC extension. Look in the folder for their exact names. You will be able to add others shapes.
Open the folder and look for the icon of name BEYCA! Click on it with the right button  of the mouse and to choose send> desktop. (Or create a connection and  move the new icon on the desktop).
Then click on BEYCA! now on the desktop. A page will appear. Write the name of the image to recognize, for example if you write A (without .PIC extension), four images of A appear, …. Center one of this four images in the square. Its contour must be far away from the perimeter of the square! Often crash of the program  are resolved moving the square.
The elaborated and codified forms go in the file RICORDI. If you delete it you erase the whole of the  memorization.
Who wanted to add images refers to the " Read-me" in the folder of the images.

Suggested method of teaching.

Behave whit the computer like you would behave whit a child. If the child does not know nothing (in our case RICORDI it is empty) you could teach thus:
1. Launch the program, clicking on the icon BEYCA!. In the screen write A so you recall the four forms A, isolate that  on top left and follow the instructions that will appear. The computer will say  you that it does not recognize the form and  will ask  its name to you. You can answer AL (ALetter or ALPHA or simply A). Close the program.
2.Launch the program, clicking on the icon BEYCA! Recall the four forms B, isolate that on top left and follow the instructions. The computer will say to you that " perhaps" it is AL, you answer nay and inserted the name of the form: BI. Close the program.
3. Idem with C.
4.Idem with D.
5. Idem with and.
6.Idem with F.
Then begin to interrogate the computer.
Launch the program. Recall the four letters A. Isolate the letter on the top right. The computer will answer  you that is AL. The name that you have given at the shape on the top left. Recall the four letters B. Isolate the letter B on the top right. The computer will answer you that is BI. Name that you have given to the shape on the top left. Interrogated for a while; then teach the name of other forms G, H, I, L, M. Teach with the form  on the top left. Then recommenced to interrogate up inserting the shape on the top right. Behave in the same way for N, O, P, Q, R, like for the rest dell' alphabet: teach the name of the form top on the left and  ask to recognize that top on the right. At the end interrogate with letters on the bottom  left and  right.
Therefore you may add other forms that you will design.
I have not found errors in the recognition. Instead I have noticed a bug that it prevents to elaborate letter M.
Reading often the notebook "Bugs and developments" is recommended.

Remarks: S= yea , N= nay; Invio= enter.


You may widely change the predefined parameters in the sources: the result will change little. This in general terms. Attention! changing SMO2 in the program RICON and the parameters and the function that defines the memorization in the program MISCHIA: type of intelligence of the machine changes. REMARK. Topological recognition is demanded generalizing and approximate in the syntactic recognition, no matter if it discriminates little between similar forms.