Planogram Matching Control in Grocery Products by Image Processing

Abstract

The placement of products in grocery shelves is designed according to several marketing strategies. These designs are represented by schemas named as planograms. Currently, the planogram matching control of grocery products are counducted by a personel manually and causes time loss. In this paper, a system that aims to solve this problem is proposed. The sytem takes a grocery shelf image as the input and gives the planogram compatibility as the output. Proposed system performs with % 99.03 precision in shelf line detection, % 97.31 precision product in detection and % 99.21 accuracy in brand identification.