Gas injection is a common practice in numerous metallurgical vessels, and continuous casting mold is one of the examples of such vessels. Argon gas injection into the submerged entry nozzle contributes to increasing the sequence length. However, in some unfavorable conditions, these bubbles become responsible for the occurrence of steel defects such as blister, sliver, and mold slag exposure, etc. Bubbles size distribution in the mold is one of the indicators of cast steel slab quality. Previous studies on estimating the bubble size distribution in mold lack in providing the challenges of automatic image processing (IP) techniques employed to measure the size of the bubbles. On the other hand, bubble size measurement studies performed on other reactors such as bubble columns, etc., are programming intensive and need in-depth knowledge of coding. Some automatic IP techniques also exist which facilitate a user-friendly environment and quick estimation of bubble characteristics. Even though these IP techniques are easy to use, an assessment of their applicability with respect to experiments performed in a particular study is required. Therefore, in this study, details of imaging and automatic IP techniques are discussed through a comprehensive comparative analysis. Three different IP techniques were examined for measuring the bubble sizes, and the performances of these techniques were compared with respect to a newly developed manual benchmark technique. One of the identified techniques, Ellipse Split, was applied to estimate the bubble size distribution at different gas fractions for validation.