Autonomous Navigation is an indispensable technology in which Line-of-Sight (LOS) Vector is very essential. Measuring LOS Vector involves image processing techniques and the existing methodology for this measurement starts with pre-processing of captured raw image of target celestial body. The preprocessing step includes edge detection by Canny Edge Detection Algorithm. Fitzgibbon's Approach to least squares fitting of ellipse is used for fitting the ellipse to the edge data points and to determine the ellipse limb points. With these limb points, ellipse parameters such as ellipse centre, semi-major axis, semi-minor axis, and the inclination angle from the x-axis to the ellipse major axis are computed. Subsequently, the LOS Vector from the spacecraft to the centroid of the target celestial body is obtained. But in this proposed work, a Modified Morphology Based Edge Detection Algorithm is used in the place of Canny Edge Detection Algorithm. Further, the line of sight vector and error in computation is analysed and we noticed that the error is minimized through the proposed method.