Algorithm for processing high definition images for food colourimetry

An algorithm was developed for image processing of High Definition (HD) images for food colourimetry applications. HD images help to capture detailed features and provide higher measurement accuracy. Dense image matrix of HD images makes image-processing complex and requires high-end data processing system. A two-step image cropping was found useful for rapid analysis of HD image and in reducing the computational power requirement. CIE Lab colourimetric parameters were obtained by transforming RGB image into CIE Lab image. Algorithm was used for image pre processing and was evaluated to reduce the image matrix size. Computing performance in terms of CPU, RAM and Disk usage was determined during the processing of HD images for colour extraction. At 1100 similar to 1100 image size, image processing was successfully completed while keeping the CPU usage within stable range. Algorithm was integrated into food colour vision system to determine CIE Lab values of skim milk powder. (C) 2020 Elsevier Ltd. All rights reserved.