Digital image segmentation is a well-researched hard problem of highlighting each object in an image with a different shade. It increasingly relies on a superpixel approach to circumvent the computational sumptuousness inherent in a high number of image pixels. Superpixel clusters similar pixels in primitive features such as colors into semantically related groups of fewer pixels to accelerate computation. However, most of the existing superpixel methods rely heavily on Euclidean distance to measure color similarity between two image pixels. Unfortunately, Euclidean distance assumes that data samples are distributed about the mean in a spherical fashion and may not compactly measure color similarity to comply with human perception of object similarity. This study was aimed at experimentally comparing Euclidean distance with strong attribute concurrence influence distance (SAID) in simple linear iterative clustering (SLIC) superpixel segmentation algorithm to determine the effects of distance measures on the performance of SLIC. In addition, the density-based spatial clustering of applications with noise (DBSCAN) was used as a control superpixel algorithm to comprehend the efficiency, ability to adhere to image boundaries, and effects of distance measures on SLIC performance. Experimental results showed that DBSCAN achieved better results in terms of under-segmentation error for regular compactness in complex, multiple, and overlapping images. However, SLIC with SAID distance achieved better results in terms of achievable segmentation accuracy, and superpixel compactness for complex images. Moreover, SLIC superpixel algorithm achieved better results with Euclidean distance for normal images with less complex backgrounds when compared with SLIC with SAID distance and DBSCAN superpixel algorithm.