The image data sets provided in (Dias et al., 2018b) consist of four datasets: AppleA, AppleB, Peach, and Pear. These images were created from different fruit flower species with different capture angles under varying natural environments, as summarized in Table 1. Both datasets AppleA and AppleB are images of apple trees, which consist of 147 images and 18 images, respectively. The Peach and Pear datasets are composed of 24 and 18 images, respectively. Labeling of the datasets combines freehand