Background and Introduction

Currently, PT Perkebunan Nusantara VIII Rancabali contributes to national tea production with a daily production of 35–40 tons of black tea. However, At PTPN VIII, the black tea quality test system still uses the traditional method of only testing by eye, so there are indications of human error. This is inefficient and does not meet consumer demand. Should any products prove unsuccessful, return them to PTPN VIII Rancabali for additional processing.

Our Solution

The solution idea uses image processing, specifically the CNN algorithm with ResNet50 architecture, in our innovative application for orthodox black tea quality test report. This technology can classify types of black tea powder based on the color and texture of the tea powder.