The invention provides an embedded part detection method and system, and belongs to the technical field of prefabricated buildings, and the method comprises the following steps: S0, obtaining the standard position coordinate and type of an embedded part; S1, photographing a prefabricated part and an embedded part to obtain an actual assembly image; s2, correcting the distortion of the actual assembled image to obtain a corrected assembled image; s3, inputting the corrected and assembled image into a deep learning classification network, and identifying position coordinates and types of the embedded parts; and S4, comparing the position coordinate and the type of the embedded part with the known standard position coordinate and the type of the embedded part. According to the embedded part detection method provided by the invention, manual measurement and detection are not needed, time and labor are saved, the detection result is accurate, a photographing mode is adopted, the detection time is saved, an actual image of the embedded part can be photographed, accurate identification of the embedded part is facilitated, a new prefabricated part does not need to establish an image standardization template, and the operation process is simplified.