| 摘要 |
<p>An image recognition method and apparatus in which a character region is segmented from input image information and the segmented character regions are recognized. Segmenting and recognition of character regions are performed in parallel with the input of image information. In response to to an interruption in input of image information, the recognition process is halted until the segmenting process is completed. Thereafter, recognition processing may be performed on any desired ones of the segmented character regions. <IMAGE></p> |
