An elevator wire rope is a key component for carrying the weight of carriages. To address the problem that wire rope defects cannot be detected online, a method to identify broken wires using hyperspectral image processing is proposed. First, the spectral band image with the greatest difference between the sample and background was selected. Then, Otsu adaptive threshold segmentation combined with the Hough transform was used. The number of pixels in the diameter of the wire rope in the image was used to determine whether the extent of breakage meets the scrapping requirement. The mean square error, peak signal-to-noise ratio, and structural similarity values experimentally obtained using this method were 650.9, 19.9957, and 0.9404, respectively. Results indicate that combining Otsu threshold segmentation and the Hough transform to process images of elevator wire ropes is effective, providing a new method for the rapid detection of wire rope defects.