Insect detection from imagery using YOLOv3-based adaptive feature fusion convolution network

Abstract

Context Insects are a major threat to crop production. They can infect, damage, and reduce agricultural yields. Accurate and fast detection of insects will help insect control. From a computer algorithm point of view, insect detection from imagery is a tiny object detection problem. Handling detection of tiny objects in large datasets is challenging due to small resolution of the insects in an image, and other nuisances such as occlusion, noise, and lack of features. Aims Our aim was to achieve a high-performance agricultural insect detector using an enhanced artificial intelligence machine learning technique. Methods We used a YOLOv3 network-based framework, which is a high performing and computationally fast object detector. We further improved the original feature pyramidal network of YOLOv3 by integrating an adaptive feature fusion module. For training the network, we first applied data augmentation techniques to regularise the dataset. Then, we trained the network using the adaptive features and optimised the hyper-parameters. Finally, we tested the proposed network on a subset dataset of the multi-class insect pest dataset Pest24, which contains 25 878 images. Key results We achieved an accuracy of 72.10%, which is superior to existing techniques, while achieving a fast detection rate of 63.8 images per second. Conclusions We compared the results with several object detection models regarding detection accuracy and processing speed. The proposed method achieved superior performance both in terms of accuracy and computational speed. Implications The proposed method demonstrates that machine learning networks can provide a foundation for developing real-time systems that can help better pest control to reduce crop damage.

Underground cable laying environment survey method based on wide-angle image distortion removal algorithm

“…In the distortion-corrected wide-angle depth image of the downhole cable laying environment, perform the cable nameplate area selection operation, and use the SIFT feature descriptor operator to perform effective feature extraction on the area. After the feature extraction is completed, input it into the YOLOv3 network structure to detect the stickers on the cable [6], and combine it with the wide-angle depth image of the cable laying environment to obtain the depth value of the cable label in the wide-angle depth image of the cable laying environment, thereby completing the arm hand-eye calibration and camera internal parameter calibration operation. Based on the camera imaging principle, solve the coordinate position of the cable nameplate in the mechanical wall coordinate system.…”

“…thereby achieving binary thresholding of the cable nameplate gray image and effectively segmenting the cable nameplate image to obtain foreground and background information of the cable nameplate image (6). Cable nameplate text recognition based on OCR.…”

Tracking the diverse pathways to value for digital agriculture

“…The openness of the public research process is seen as vital for the development of next-generation phenotyping (Tripodi et al 2023). It is also required for the development of advanced weed or insect recognition cited in Australia (Amrani et al 2023;Mahmudul Hasan et al 2023), although in all such cases, the investment needed to scale these applications up will almost inevitably move the IP into the domains of the third classspecialist supplierbefore digital agriculture applications can be offered to farmers. Jackson and Cook (2023) illustrate the opportunities and challenges of pulling digital technologies into the livestock industry in Australia.…”

“…2023). It is also required for the development of advanced weed or insect recognition cited in Australia (Amrani et al . 2023; Mahmudul Hasan et al .…”

Feature Refinement Method Based on the Two-Stage Detection Framework for Similar Pest Detection in the Field

“…Many researchers [26,27] constructed databases by fixed pest collection devices and utilized detectors such as YOLOv5 for pest detection. In order to enhance the discrimination ability of detectors for multiple categories of pests, feature fusion [28,29] was considered for algorithmic improvement, and it was experimentally demonstrated that feature fusion is effective in improving detection accuracy. Classification-based methods [21][22][23] focused on the global features of images.…”