Surface Mount Technology (SMT) is now widely used in the electronics assembly industry. In the SMT process, Automated Optical Inspection (AOI) technology based on image processing is widespread used for chip component defect detection. However, it is plagued by several challenges including slow response time and a high false positive rate, warranting the need for further research and advancement. In recent years, researchers have turned to deep learning-based object detection algorithms for industrial defect detection. Nevertheless, in the SMT scenario, there are challenges such as complex object shapes and the difficulty in balancing accuracy and speed. Common object detection algorithms often fail to meet the requirements in such cases. To solve these problems, this paper proposes a novel real-time detection method for SMT chip component defects based on Detection Transformer (SMT-DETR), which has good results in detecting chip components with deformation characteristics, and facilitates the detection of small targets with fast convergence speeds. Experiments show that the proposed method is better compared to classical object detection algorithms.