QiandaoEar22

QiandaoEar22 is a high quality noise dataset for identifying specific ship from multiple underwater acoustic targets using ship radiated noise. This dataset encompasses 9 hours and 28 minutes of real-world ship radiated noise data and 21 hours and 58 minutes of background noise data. Collected using digitalHyd SR-1 self-capacitance hydrophones in a real underwater acoustic environment at Qiandao Lake, China, during June 2022, this dataset serves as a high-quality ship radiated noise recording. The experimental period coincided with the tourist season, resulting in recorded ship radiated noise primarily originating from sightseeing boats and speedboats. Due to the hydrophone's proximity to the navigation channel, passing vessels, including sightseeing boats, speedboats, and other watercraft, contribute to a composite radiated noise. Additional targets include large tour boats, cargo ships, sanitation vessels, small boats, research ships, and ambient noise. Moreover, mechanism of collection of data and its labeling are also discussed. Utilizing the QiandaoEar22 dataset, we built some sub-dataset which help us to distinguish ship signals and background noise signals from the underwater acoustic data and successfully recognizes three specific types of ship signals—speedboats, KaiYuan, and UUV—from the multi-target ship signals.