2021 IEEE 5th Information Technology,Networking,Electronic and Automation Control Conference (ITNEC) · 2021 · 13 citations · 12 references
ComputingComputer SciencePlant PathologyComputer VisionMachine LearningMachine VisionAgricultural Machine VisionDeep LearningImage Feature EngineeringObject DetectionObject RecognitionRepresentation LearningCluster AnalysisConvolutional Neural Network DetectionConvolutional Neural NetworksData AugmentationDeep Learning DetectionDetectorsFeature FusionReal-time Processing
In this paper, we proposed an object detection algorithm based on deep learning for NVIDIA Jetson Xavier NX to detect cucumber leaf diseases in real time. We collected cucumber leaves images from the experimental base of Anhui Academy of Agricultural Sciences, and labeled dataset by VGG Image Annotatort. Based on the one-stage detection network YOLO v5, firstly, the size of anchors were obtained by cluster analysis of the dataset, and a variety of data augmentation was carried out for to enrich the background information, The feature fusion network is improved, and different scale features are given different weights. Based on these optimizations, our model achieved 84.6 mAP on our dataset with 34.7 MB parameters, 4.9 ms inference time on NVIDIA Geforce GTX1080Ti, and 23 FPS with 512×640 resolution on NVIDIA Jetson Xavier NX, being faster than previous detectors, basically met the purpose of real-time.
12
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Ross Girshick, Jeff Donahue, Trevor Darrell et al. · 2014 · 31.8K citations
Focal Loss for Dense Object Detection
Tsung-Yi Lin, Priya Goyal, Ross Girshick et al. · 2017 · 25.7K citations
Path Aggregation Network for Instance Segmentation
Shu Liu, Lu Qi, Haifang Qin et al. · 2018 · 8.8K citations
EfficientDet: Scalable and Efficient Object Detection
Mingxing Tan, Ruoming Pang, Quoc V. Le · 2020 · 8.4K citations