Quick response (QR) codes are two-dimensional (2D) barcodes that are widely used in food packaging. Moreover, scanning QR codes with mobile phones has become a convenient method to ensure product traceability. However, attaching QR codes to curved agro-food products to improve traceability effects and increase customer satisfaction presents a significant challenge. This paper reports an optimization method developed for curved surfaces that uses response surface methodology (RSM) analysis. The value ranges of the following four factors affecting QR readability were selected and analyzed: reading distance, ball diameter, code size, and coded characters. A central composite inscribed experiment using four factors with five levels was designed using RSM to obtain the optimal reading parameters, and experimental equipment was designed in-house. The results indicate the primary factors that significantly affect QR code readability are reading distance, ball diameter, the interactions between reading distance and code size, and the interactions between ball diameter and code size. Optimal parameters were obtained with the established model by using apples and melons. For actual experimental testing, two solutions for apples and one solution for melon were selected to determine the requisite average values for the parameters. Tests were performed using different mobile phones, including both iOS and Android platforms. An analysis of the results reveals only slight differences between simulated readability and actual readability for different fruits and mobile phone platforms.
