Dark pool platforms are commonly utilized for executing large block transactions, with users paying higher fees in exchange for the privileges afforded by these platforms. In such instances, users have the capability to conduct transactions within the dark pool, concealing their trading intentions until transactions are packaged into blocks. Given the confidentiality advantages offered by dark pools, a detailed analysis of the specific behaviors and characteristics of dark pool transactions is highly valuable. However, investigative research has not systematically analyzed the internal trading behaviors and characteristics within dark pools. In this paper, we collected one week’s worth of transaction data, proposed a method for acquiring dark pool transactions, and demonstrated the reliability of this method. Furthermore, we conducted an analysis of the obtained dark pool transactions, revealing the presence of block trades, with 90.29% being contract invocation transactions categorized into eight types, while the remainder consisted of ordinary transfer transactions. Our analytical findings provide scientifically significant insights into the behaviors, characteristics, and purposes of dark pool transactions. Based on our results, users can better evaluate the specific behaviors of dark pool transactions.