A MASK-RCNN Based Approach Using Scale Invariant Feature Transform Key points for Object Detection from Uniform Background Scene. (2019)