Actually, text-based image retrieval is a method to retrieve the user's interested images semantically, but there still exist some problems in it such as high-priced manual annotation cost. To avoid the problems in text-based image retrieval, a considerable number of studies have been made on Content-Based Image Retrieval called CBIR over the past few years. Most past studies for CBIR focused on how to search the images most relevant to the query image without considering the concepts hidden. However, CBIR systems encounter the problem of high computation cost due to high visual feature dimensions. To cope with the problems, in this paper, we propose a Pattern-Based Conceptual Image Retrieval method named PBCIR to convert visual features into visual patterns. By visual pattern matching, the relevant images and concepts can be derived to achieve the purpose of semantic image retrieval. The experimental results show that, the proposed method can capture the user's visual and conceptual intents more effectively and efficiently than that considering only visual features.