Study on Apple Texture Measuring Equipment Manipulated with Hand

This paper describes a manual, compact device for measuring apple texture. Generally, in texture-measuring devices, the probes and plungers are automatically controlled at a fixed speed and load. However, this paper aims to develop simple and portable equipment in which the human moves the probe to stab the apple flesh with their hand. In the experimental part, we measured the texture of three brands of apples using the proposed equipment. The analysis revealed that the textures of the three apple brands differed. Still, the authors reasoned it was challenging to distinguish between those textures due to the considerable variation in the same brands’ apple data. Therefore, we decided to adopt a neural network model that can classify apple brands. As a result, we confirmed that the classification accuracy mean was more than 90% using the proposed neural network model.