Option B is the correct choice because Cloud Functions can directly respond to changes in a Cloud Storage bucket, such as new object uploads, by triggering the function automatically. This method is efficient and specifically designed for such use cases. Option A suggests using Google Kubernetes Engine, which is overkill for this simple task. Option C involves App Engine and Cloud Scheduler, which doesn't directly address the need for triggering on uploads. Option D proposes Dataflow, which is more suited for large-scale data processing rather than immediate, small-scale tasks like resizing images upon upload. For more details, refer to Cloud Functions documentation.