Based on attachment position information of each of a plurality of sensor units, which reads an image in each area obtained by dividing the area of an original in a sub-scanning direction, image data output from the sensor unit undergoes attachment error correction of correcting the attachment error of the sensor unit. Image data output from the plurality of sensor units then undergo joint processing. Based on the placement information of the original, tilt correction of correcting the tilt of the original is performed for the image data having undergone the joint processing.