Evaluation of fruit selection with ensemble model in the Peruvian industry

This paper allows to evaluate fruits with the ensemble algorithm with artificial vision, besides, it evaluates the systematic review with the PRISMA methodology thought the last 10 years. Our findings are the application of TensorFlow and self-supervised multi-network fusion classification models. Some researchers describe classification errors. The deep learning methodology is also an efficient way of classifying and applying algorithms that optimally analyze features such as color, shape, and texture of the object. On the other hand, we can implement artificial intelligence to this work since in recent years it has been developing in all industrial sectors and is bringing positive results. Finally, it uses a case study with the limitations to admit safe practices in a company dedicated to agricultural production and also over time that when studying the last cycle and at the same time working, the availability of time is limited to be able to use it in a work area. The results are an mean absolute error for size evaluation of 92.92%, color evaluation with 93.23%, and the assessment of the maturity, defects and quality of 97.72% after the evaluation of 1271 research articles with the systematic review process.