PAL-DET

The dataset used in the experiment consisted of four datasets: "Package and label detection", "Cardboard3 Computer Vision Project", "Packages on Conveyor", and "0628huashu_gongbaotai_product2". We removed some of the poorer quality images from the original "Package and label detection" dataset and added images from the courier sorting site and some images from the "Packages on Conveyor", "Cardboard3 Computer Vision Project" and "0628huashu_gongbaotai_product2". For these new images, we used the Make Sense tool to re-label and augment the data. The final fused PAL-DET dataset contains 13,047 images, which are divided into two categories: package and label. In order to ensure the generalization ability of the model in practical applications, we divide the dataset into a training set, a validation set, and a test set at a ratio of 8:1:1.