The detection of product defects are crucial in internal control in manufacturing. This study surveys state of the art deep-learning methods in defect detection. First, we classify the defects of products, like bottles, toothbrushes, leather, capsules, hazelnut, screws into categories. Next, recent mainstream techniques and deep-learning methods for defects are reviewed with their shapes and sizes described. Next, we summarize and analyze the application of deep learning, machine vision, and other technologies used for defect detection, that specializes in these aspects, namely method and experimental results. To further understand the difficulties within the field of defect detection, we investigate the functions and characteristics of existing equipment used for defect detection. The core ideas and codes of studies associated with high precision, high positioning, rapid detection, small object, complex background, occluded object detection and object association, are summarized. Lastly, we outline the achievements and limitations of the prevailing methods, together with the present research challenges, to help the research community on defect detection in setting an extra agenda for future studies.