Research On The Detection Method Of Logistics Package Considering Different Regions

As an information carrier with rich semantics,image plays an increasingly important role in real-time monitoring of logistics management.Abnormal objects are typically closely related to the specific region.Detecting abnormal objects in the specific region is conducive to improving the accuracy of detection and analysis,thereby improving the level of logistics management.In the field of logistics and transportation,due to the large number of express delivery items,it is common to quickly sort goods,resulting in the loss of items.In the process of loading and unloading,the soft and uneven goods at the bottom of the package will drop when they are transported on the conveyor belt,and the phenomenon that the goods are forgotten under the conveyor belt often occurs due to the busy work of the workers.However,at present,the object detection method is mainly used in the logistics management to identify the goods in the logistics transportation,which can not further distinguish the goods,and can not detect the abnormal objects in the goods.Based on these observation results,we first get the data set needed in this paper by preprocessing the actual pictures of the warehouse.And the data set format obtained by labeling with labelme is similar to coco data set,and saved in JSON format;Then we design the method called abnormal object detection in specific region based on Mask R-CNN: AODin SR.The first step of this method is to use the traditional Mask R-CNN method to get the initial instance segmentation model.The second step is to calculate the region overlap of the specific region and determine the overlapping ratio of each instance,and combine these two parts of information are fused to predict the exceptional object.The third step is to recover and detect the abnormal object in the original image.Finally,the experimental results on real data set show that AODin SR can effectively identify abnormal objects in a specific region,in order to steadily increase the effect of exceptional objects detection.