After the rapid urbanization in the last several decades, the land urbanization, which is characterized by urban expansion, has gradually become saturated. Land use strategies have gone through the stage of "inventory planning" in many cities of China. At this stage, urban renewal programs can intensify dynamics of urban land cover, which leads to change the local landscape pattern, and finally has effects on different ecological processes and functions. However, the existing studies mainly focus on urban external expansion and lack understanding of dynamics within the city. In our study, we identified three types of land development modes: urban expansion, urban renewal, and infilling development in Shenzhen based on high-spatial resolution remote sensing imagery. The landscape pattern changes in different modes were compared and analyzed on changing parcel scale. The results showed that: (1) Shenzhen had entered the stage of "inventory optimization" in urban development, showing that the area of internal inventory changes was twice that of external incremental changes. (2) The average proportion of vegetation in urban renewal types had increased by 3%, while that of urban expansion and infilling development had decreased by 51% and 50%, respectively. (3) Urban expansion and urban renewal had obvious spatial aggregation characteristics, and the distribution of infilling development was relatively random. This study identified three different types of landscape changes and their spatial distribution characteristics on a fine spatial scale. It provided a new insight for a deeper understanding of the dynamic changes in landscape pattern within the city, and it would provide a support for the mechanism study of urban ecological process and urban planning.