In the rapidly changing environment of warehouse automation, efficient management of piles of objects in unordered, random arrangements remains a formidable challenge. The paper addresses this challenge with a novel approach to detect the geometric features such as edges and corners in the unordered 3D point clouds tailored for robotic operations. The proposed method employs an eigenvalue based surface variation measure to rapidly extract sharp edge points from raw point cloud data, offering improved speed and efficiency compared to traditional approaches. Additionally, a 3D Harris corner detector is also used to identify prominent corner points that subsequently form the foundation of trustworthy pose estimation of texture less objects. When used with synthetic shapes, the technique achieves unprecedented effectiveness in delivering fast and accurate results. It takes much less computation time as compared to the previously reported algorithms. This makes it an efficient transformative tool for real time pick and place tasks. This advancement helps autonomous grasping in cluttered warehouse settings, allowing for more intelligent and efficient automation in the building and manufacturing industries.