Foreign matter present inside an object and the like are detected with higher resolution and higher reliability. Frame data of a plurality of tomographic planes that are parallel in the scan direction and set in a space between an X-ray tube (31 and an X-ray detecting unit (22) is generated based on detected frame data. The generation of frame data is performed based on the fan-shaped spreading of an X-ray beam and the differences in position in a height direction between the plurality of tomographic planes from a detection surface. Tomographic images are respectively generated from the frame data of the plurality of tomographic planes based on a laminography technique. Edge information based on the changes in pixel values in each tomographic image is calculated for each pixel. A three-dimensional distribution of the edge information is generated. A search is performed on the edge information in a direction passing through the plurality of tomographic planes. As a result of the search, pixels indicating a maximum value in the edge information are detected. Only pixels in the plurality of tomographic images that positionally correspond to the detected pixels are combined into a single composite image.