CNN-ENABLED DETECTION SYSTEM FOR AGRICULTURAL ANOMALY ANTICIPATION
Artificial intelligence is a rapidly developing field today. One of its different applications is object acknowledgment, utilizing PC vision. The advancements in Deep Learning (DL) techniques have made it possible to quickly identify, localize, and recognize articles from images or recordings. A gro...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Kragujevac
2024-12-01
|
| Series: | Proceedings on Engineering Sciences |
| Subjects: | |
| Online Access: | https://pesjournal.net/journal/v6-n4/57.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Artificial intelligence is a rapidly developing field today. One of its different applications is object acknowledgment, utilizing PC vision. The advancements in Deep Learning (DL) techniques have made it possible to quickly identify, localize, and recognize articles from images or recordings. A growing number of farming and agricultural applications are utilizing deep learning techniques. Automatic weed recognition and grading can assist with weed control and thereby contribute to higher yields. Weed identification in crops through symbolic means is inherently problematic due to the presence of comparable varieties (’green-on-green’) between the two weeds and harvests, as well as the similarity in their developmental morphologies and surfaces. The mix of these two advances prompts the justification for this undertaking. In this proposed model, the essential point is to perceive the weeds in crops by using YOLOv3 with PyTorch and OpenCV. The presentation of the subsequent framework is contrasted, and comparable ventures are found during the analysis. A technique is created to gather information for weed recognition, along with a pipeline to deal with the pictures. The information will be utilized to prepare cutting-edge object identification models like YOLO. To find an ideal model for the continuous discovery of weeds, the model will foster an information assortment procedure that will be utilized in a CNN and profound learning base model to anticipate yields and weeds uniquely in contrast to symbolism. This work centres around weed recognition using picture-handling strategies. |
|---|---|
| ISSN: | 2620-2832 2683-4111 |