We introduce a novel shape-based feature set, termed the Histograms of Categorized Shapes (HCS), for robust Three-Dimensional (3D) object recognition. By adopting the sliding window approach and a linear Support Vector Machine (SVM) classifier, the efficacy of the HCS feature is assessed on a 3D ear detection task. Experimental results demonstrate that the approach achieves a perfect detection rate, i.e., a 100% detection rate with a 0% false positive rate, on a validation set consisting of 142 range profile images from the University of Notre Dame (UND) 3D ear biometrics database. It is to the best of our knowledge that the detection rate achieved here outperforms those reported in the literature for the given dataset. The proposed detector is also extremely efficient in both training and detection due to the simplicity of the feature extraction and speed of the classification process, suggesting that the method is suitable for practical use in 3D ear biometric applications.