Algorithms evaluation for improving classification and counting model in warehouse receiving management

The purpose of this research is to evaluate algorithms and select a suitable one to improve the data accuracy of industrial components classification as well as counting in the warehouse receiving management. The modified algorithm is based on the colour space principle. We performed a series of experiments by adjusting the ratio of the white colour histogram to receive a superior performance for the proposed ANNCIC model (Artificial Neural Network Model for Components Identification and Counting). The tasks in this study consist of industrial images collection, pre-processing of the image data set and experiments. The outcome of experiments demonstrated the histogram correlation coefficient of the white colour has outperformed the pixel standard deviation which has achieved an accuracy of 93.75 per cent in classification and a 94.29 per cent accuracy in counting. This improved approach is worth more investigation by an extensive image data set of industrial components in the receiving management.