Character recognition in input images divided into areas

An image processing method comprising:

• a step of photoelectrically converting a document image which has been optically read and inputting the converted document image as image data;
• a step of separating a plurality areas included in the input image data from one another;
• a step of performing a character recognition from a plurality of directions with respect to character data in the input image data with reference to a character recognition dictionary in which features of characters have been stored;
• a first discriminating step of discriminating directional reliabilities based on results of the character recognition per direction of a plurality of characters extracted from each area;
• a second discriminating step of discriminating a directional reliability for an entire area of the input image data, based on the discriminated directional reliabilities for each area;
• and
• a third discriminating step of discriminating a document direction based on the directional reliability for the entire area.

A computer-readable medium storing instructions executable by a processor-controlled image processing apparatus, said instructions causing the apparatus to perform the following steps:

• a step of photoelectrically converting a document image which has been optically read and inputting the converted document image as image data;
• a step of separating a plurality areas included in the input image data from one another;
• a step of performing a character recognition from a plurality of directions with respect to character data in the input image data with reference to a character recognition dictionary in which features of characters have been stored;
• a first discriminating step of discriminating, directional reliabilities of each of the plurality of areas based on results of the character recognition per direction of a plurality of characters extracted from each area;
• a second discriminating step of discriminating a directional reliability for an entire area of the input image data, based on the discriminated directional reliabilities for each area;
• and
• a third discriminating step of discriminating a document direction based on the directional reliability for the entire area.