This paper presents a comparative analysis of traditional edge detector operator with the proposed algorithm on grayscale images. The proposed algorithm is based on mathematical morphology and thresholding. Mathematical morphology is a new technique for edge detection based on set theory. It has been used for feature extraction and feature detection. Basic operations of Mathematical morphology are dilation, erosion, opening and closing. Based on these operations, experiment results are obtained using square structuring elements of different sizes with different images. The authors adopt the thresholding to change the brightness of edges of image. Detection of edge is a preprocessing step in image processing. Edge detection has been done using traditional operators (Canny, LoG, Prewitt and Sobel). The edge detection process had been used to reduce the amount of data and filter out useless information. The aim of this paper is to obtain the useful edges of the image object. This paper prefer round shape object images to extract the edges using different combination of structuring elements. Performance evaluation of the proposed algorithm is based on parameters like Root Mean Square Error (RMSE). This parameter is used to calculate the image quality of an output image. The experiment result shows that the proposed algorithm has more superior results than traditional operators.