摘要 |
PROBLEM TO BE SOLVED: To provide an image processor by which operation becomes simple and an image can be trimmed precisely and outputted, by avoiding requiring exactness in the designation of a trimming area. SOLUTION: The inside of a designated trimming area R and the peripheral part of the area are divided into plural blocks depending on the state of a ruled line L and a blank part W, and the character recognition of these blocks is performed. Furthermore, concerning a block partially included in the trimming area, context recognition between with a block entirely included in the trimming area is performed to detect the connection of sentences to discriminate whether an image is within the trimming area. Thus, an image V in a desired area R can be trimmed precisely and obtained without detailed trimming instruction.
|