Abstract: Taking Harbin city as the research object, the paper extracted the vegetation cover based on Landsat TM/OLI remote sensing images in 2004, 2010, 2015 and 2020, which combined the normalized vegetation index and dimidiate pixel model. Then, we integrated the transfer matrix to analyze the spatial and temporal dynamic changes of vegetation coverage. Based on the results of vegetation coverage classification, four landscape pattern indices namely patch area, number of patches, contagion index, and Shannon’s diversity index, were calculated by Fragstats4.2 software to study the spatial and temporal evolution characteristics of the landscape pattern in Harbin. The results showed that the sum of areas with high vegetation coverage and medium-high vegetation cover accounted for 77.81%, 72.54%, 73.01% and 83.31% of the total area in 2004—2020, respectively, and the vegetation coverage exhibited an overall increasing trend in Harbin City. The landscape percentage of high vegetation coverage was the highest from 2004 to 2020, followed by medium-high vegetation coverage, and the number of patches exhibited a first increasing and then decreasing trend for the most vegetation coverage region. At the landscape level, the aggregation of patches decreased and the landscape diversity increased slightly. This study explores the dynamic changes of vegetation coverage and landscape patterns in Harbin city, which can provide guidance and reference for the planning design and landscape pattern optimization of the city.