Four-Dimension Deep Learning Method for Flower Quality Grading with Depth Information

Grading the quality of fresh cut flowers is an important practice in flower industry. Based on maturing status, a classification method based deep learning and depth information was proposed for grading quality. Firstly, RGB image bud were collected transformed into fused RGBD information. Then, set as inputs convolutional neural network to determine status. Four models (VGG16, ResNet18, MobileNetV2, InceptionV3) adjusted four-dimensional (4D) input classify flowers, their performances compared with without The experimental results show that accuracy improved information, InceptionV3 achieved highest (up 98%), which means can effectively reflect characteristics helpful These have certain significance intelligent sorting flowers.