Urban green spaces aim to provide citizens with calm environments free of annoying city noises. This requires a thorough understanding of noise emission and related exposure to sounds in green spaces. This paper investigates noise perception in several different spots of a specific urban green space in the city of Dresden, Germany. The measurements were conducted in a large park called Großer Garten, which covers 1.8 km² of various landscapes, including streams, park railways, fountains, paths for bicycles and pedestrians, etc. In this research, noise perception was investigated at eleven spots that were chosen in a way to reflect different noise characteristics of the soundscape. Specifically, the exposure spots emphasize four sound types: nature, human-induced, traffic, and mechanical sounds. In parallel, binaural audio and video recordings were conducted for each spot to identify the connection between the perceptual measures of the specific location and its respective psychoacoustic parameter values. Subsequently, an annoyance response for each spot was identified based on the corresponding participants' perception and was associated with the relevant parameters. Moreover, the evaluation spots were categorized based on psychoacoustic parameters. Eventually, correlations between four investigated types of noise, annoyance, and phychoacoustic parameters were derived and interpreted.