Industrial image processing and computer vision play significant role in factory automation since industries now employ human-robot interaction for the monitoring of products in areas considered risky and dangerous for humans. The challenge however, is to ensure reliability in image processing such that image sizes and image similarity index are expected to be perfect representation of actual objects. This paper investigated the statistical relationship between the image ratio size and similarity index after compression. Data set collected from the simulation of a selected image was analyzed and compared with data sets from two similar works. Using correlation analysis, a statistical relationship was established between the image size ratio and similarity index of selected images under review. It was observed that an inverse and high correlation existed between the image similarity index and image size ratio of compressed images. The result of the validation shows that the proposed regression model has predictability or good fit of 95%.