PURPOSE: To efficiently and surely execute desired recognizing processing by providing a recognizing part which is provided with the neural layer of layer structure to meet the degree of abstractness of the fine lined picture element data and recognizes a graphic according to the thinned picture element data.
CONSTITUTION: A picture processing part 300 converts the graphic, into a fine lined picture and sends the picture element data to the recognizing part 100. The recognizing part 100 is constituted of a neural network to execute graphics recognition based on the fine lined picture element data, and is provided with the neural layers 20, 90. Then, the data of low order among the fine lined picture element data is inputted to the neurons 21 to 23 of the layer 20, and besides, the data of the high degree of abstractness such as a number of end points, the coordinates of the end point, a number of branches, a number of branch points, the coordinates of the branch point, curvature, and an inflection point, etc., is inputted to the neurons 91 to 93 of the layer 90, and they are processed. Thus, since the recognizing part 100 is of the layer structure and of data input configuration to meet the degree of abstractness of the data, the desired recognizing processing can be surely executed.
YAMAMOTO MAKOTO
MATSUMOTO KOJI
KUMAGAI RYOHEI
