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 × 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.

开发了一种算法，用于食品比色法的高清（HD）图像的图像处理。高清图像有助于捕获详细功能并提供更高的测量精度。HD图像的密集图像矩阵使图像处理变得复杂，并且需要高端数据处理系统。发现两步图像裁剪可用于快速分析高清图像并减少计算能力要求。通过将RGB图像转换为CIE Lab图像来获得CIE Lab比色参数。该算法用于图像预处理，并进行了评估以减小图像矩阵的大小。在处理高清图像以进行颜色提取期间，确定了基于CPU，RAM和磁盘使用情况的计算性能。在1100×1100图像尺寸下，图像处理成功完成，同时将CPU使用率保持在稳定范围内。该算法已集成到食品色觉系统中，以确定脱脂奶粉的CIE Lab值。