Noninvasive diagnosis of internal traits in fruit crops is a high unmet need; however it generally requires time, costs, and special methods or facilities. Recent progress in deep neural network (or deep learning) techniques would allow easy, but highly accurate diagnosis with single RGB images, and the latest applications enable visualization of "the reasons for each diagnosis" by backpropagation of neural networks. Here, we propose an application of deep learning for image diagnosis on the classification of internal fruit traits, in this case seedlessness, in persimmon fruit (Diospyros kaki). We examined the classification of seedlessness in persimmon fruit by using four convolutional neural networks (CNN) models with various layer structures. With only 599 pictures of 'Fuyu' persimmon fruit from the fruit apex side, the neural networks successfully made a binary classification of seedless and seeded fruits with up to 85% accuracy. Among the four CNN models, the VGG16 model with the simplest layer structure showed the highest classification accuracy of 89%. Prediction values for the binary classification of seeded fruits were significantly increased in proportion to seed numbers in all four CNN models. Furthermore, explainable AI methods, such as Gradient-weighted Class Activation Mapping (Grad-CAM) and Guided Grad-CAM, allowed visualization of the parts and patterns contributing to the diagnosis. The results indicated that finer positions surrounding the apex, which correspond to hypothetical bulges derived from seeds, are an index for seeded fruits. These results suggest the novel potential of deep learning for noninvasive diagnosis of fruit internal traits using simple RGB images and also provide novel insights into previously unrecognized features of seeded/seedless fruits.

No abstract

Crispness of persimmon flesh was evaluated by a swing-arm texture device. A conical probe was inserted into the flesh, and horizontal and vertical vibrations of the probe were separately monitored from 0 to 51,200 Hz. The vibrations were used to calculate the energy texture index. Out of the seven persimmon varieties tested, sensory tests showed that three varieties: 'Taishuu', 'Neo sweet', and 'Reigyoku', were more crisp than the other varieties: 'Maekawa-Jiro', 'Matsumotowase-Fuyu', 'Shinsyuu', and 'Kanshu'. There was a significant positive correlation between the energy texture index of horizontal vibrations and the sensory score. However, the sensory score was found to be influenced by the flesh firmness. Vibration ratios were obtained by dividing the horizontal by vertical vibrations, which were correlated with the elasticity index and sensory score. The vibration ratios calculated in 3,200-4,480 or 4,480-6,400-Hz frequency bands were especially useful for evaluating the crisp texture of each variety. In addition, the food friction index obtained by the swing-arm device had a highly significant positive correlation with the sensory score for flesh firmness and also with the elasticity index. The food friction index may be used for evaluating the flesh firmness of persimmon. These results show that the swing-arm device for texture evaluation can simultaneously measure both crisp texture and flesh firmness with single probe penetration. Furthermore, a scatter plot of the food friction index on the horizontal axis and vibration ratio on the vertical axis makes it possible to classify the flesh texture characteristics of persimmon varieties.

A storage technique for 'Taishuu' persimmon fruit was investigated using a polyethylene package, controlled O 2 and CO 2 , and a low temperature to extend the storage period so they could be given as a year-end gift, as per the traditional Japanese custom. Packing in a 0.06-mm-thick polyethylene bag preserved flesh firmness 51 days after the start of storage under 0°C, but its taste deteriorated due to an off-odor. Packing with a microperforated film highly permeable to O 2 did not suppress the off-odor, while the presence of CO 2 absorbent in the bag suppressed off-odor production and maintained flesh crispness. An increase in the CO 2 concentration in the bag caused ethanol production in the flesh even under 20% O 2 , suggesting that the accumulated CO 2 in the bag results in off-odor production. The combination of low temperature and CO 2 absorbent in the 0.06-mm polyethylene bag was proposed to extend the storage period for 'Taishuu'.