Plant root architecture plays a vital role in establishing its interaction and, thereby, performance in ground improvement. Root image analysis has been widely adopted to compute the root traits, but it has associated limitations related to traits correlation. Hence, the current study aims to develop a methodology such that constraints on adapting imaging tools for their application to root phenotyping could be grouped to eliminate impacts on root trait correlation. The root traits of five different plant species were investigated, employing two conventional computational image analysis tools to estimate root area ratio (RAR) and root shape classification. In addition, a novel set of RAR has been defined based on input parameters, i.e., length and diameter (1-D) as RAR_l,d, and volume (3-D) as RAR_v, and these measurement approaches were compared. The relationship between the measured and computed features of standard images was determined. The variations in factor values from computational tools and their reflection on the computation of RAR_l,d and RAR_v were computed. The overall RAR profile of species decreases with depth and lies in the range in accordance with other studies. Based on the RAR profile with depth, the applications of plant roots for ground improvement can be elaborated.