In the field of computer vision, instance segmentation is recognized as a crucial problem that plays a vital role in many applications such as object recognition, motion tracking, and scene analysis. However, existing methods often face difficulties when confronted with complex cases such as overlapping objects or irregular shapes. The problem of accurately separating individual objects in an image, particularly in complex scenarios, is still considered a major challenge in the research community. To address these challenges, a new method combining energy functions and points of interest is proposed in this study. The energy function is utilized to model object and background features, while points of interest are applied to guide the segmentation process. This method has been tested on single street images for visual observation and image segmentation based on the standard Cityscapes dataset for training, showing effectiveness in segmenting individual objects in images.