In a circumstance where an image processing apparatus is connected to and capable of communicating with a plurality of processing servers each performing a specific data processing service, what kind of processing is performed on document image data read out from a document by image reading means is determined in accordance with the document image data. Then an address of a processing server capable of performing the processing thus determined is searched. Then at least a part of the document image data or character-string image extracted therefrom is supplied to the address thus searched, and the data processing service is requested. From this address, a result of the data processing service is obtained, and the obtained result of the data processing service is outputted. With this arrangement, when an image is read out from a document and a result of image processing corresponding to a character-string image in the document image, it is possible to efficiently utilize the processing resources such as dictionary information and a program, which are required for performing data processing on the character-string image.