The methodology of digital image analysis has been applied to the quality determination of agricultural products from organic, conventional and hydroponic cultivation systems, as well as other food products. However, there is a very small number of studies which consider the application of this technology to fruit flours. This work aimed to characterize the instrumental color of commercial fruit flours and to correlate these parameters to the digital image processed by a specific software. The experimental design consisted of 10 different commercial fruit flours: açai (AÇ), plum (AM), green banana (BV), coconut (CO), orange (LA), lemon (LI), apple (MA), papaya (MM), passion fruit (MR) and grape (UV), each one in three different batches. The instrumental color in the CIE range (L*, a*, b*, Cab, Hab) was determined in triplicate and the images were obtained by digital camera, according to the procedure described by De Paula Filho (2013). These images were processed in a specific software according to programming defined for the study of digital images classification. Considering the color characteristics obtained for the different color parameters (L*, a*, b*; H, S, V; L, U, V and X, Y, Z) for each flour assessed in this study, data showed that these characteristics are intrinsic properties of the original fruit. An average color system was established and indicated a sample uniformity in the different batches of flour. Also, the Artificial Neural Network (ANN) analysis with Multilayer Perceptron Algorithm was performed on data set of different color parameters and showed 100 % of correct points to classify all the commercial fruit flours. ANN can be applied to fruit flour classification and this algorithm from machine learning can be used for quality control in food industries.