For any city to become sustainable, the decryption of urban growth dynamics acts as a pivotal layer. Not only from the planning perspective, resource management as well as the conservation of environmental parameters are also the primitive foundations. This research takes into account the machine learning techniques to decipher the urban growth dynamics by understanding the sprawl as well as the prediction of what the future scenarios would look like. The two techniques incorporated were the pixel-based approach and the geographic object-analysis based approach (GEOBIA) for monitoring and analyzing the urban contexts. The research also compared these two to have a clear-crisp idea about which technique is suitable, making it a contextually intelligent study. Land use land cover (LULC) was analyzed through a time-series approach between 2017 and 2023. The pixel-based technique focused on spectral signatures while GEOBIA focused on spatial as well as spectral information making a different schema of approaching the datasets. Temporal LULC variation (trends) were analyzed, quantified as well as portrayed the highlight-transitions amongst the different class-layers in the LULC dataset. The analyzed results portrayed the pros and cons of both the techniques delineating that pixel-based was suitable for fine-grained analysis while the GEOBIA approach complied more towards enhancement and noise reduction principles. Support vector machine (SVM) related algorithm was employed to granulate the two incorporated approaches. The research work baselines an efficient and effective foundation for utilizing high-resolution remote sensing data and integrate machine learning approach creating a profound layer for urban planning as well as policy-making in such treacherous dynamic environments.