Packaging is one of the most important aspects in the food industry. The problems faced during packaging can classified into two categories, defects in the packaging before the substrate is filled and after the substrate is filled. There are methods to determine the defects in the food containers after the food has been packed by means of measuring the phenomenon caused by the defects such as use of moisture sensor for detecting any leaking or the use of pH indicators. But these methods, although cost effective, slows down the plant as well as increases the risk of damaging the system. In this paper, we have implemented a defect detection and rectification system using image processing on tin cans to check for any irregularities apart from the ones that are intended to be on it. The decision-making capability is provided by a classifier neural network. The case study takes up the problems faced in coconut oil industry. A similar set up can be used after the package to detect if any defect is present. For other fluids such as soft drinks and other drinks a pH indicator can be used for detecting defects. © 2019 IEEE