Accuracy and Efficiency Comparison of Object Detection Open-Source Models

Brahim Jabir,Noureddine Falih,Khalid Rahmani

doi:10.3991/ijoe.v17i05.21833

Brahim Jabir, Noureddine Falih + Show 1 more

Open Access

https://doi.org/10.3991/ijoe.v17i05.21833

Copy DOI

Abstract

In agriculture, weeds cause direct damage to the crop, and it primarily affects the crop yield potential. Manual and mechanical weeding methods consume a lot of energy and time and do not give efficient results. Chemical weed control is still the best way to control weeds. However, the widespread and large-scale use of herbicides is harmful to the environment. Our study's objective is to propose an efficient model for a smart system to detect weeds in crops in real-time using computer vision. Our experiment dataset contains images of two different weed species well known in our region strained in this region with a temperate climate. The first is the Phalaris Paradoxa. The second is Convolvulus, manually captured with a professional camera from fields under different lighting conditions (from morning to afternoon in sunny and cloudy weather). The detection of weed and crop has experimented with four recent pre-configured open-source computer vision models for object detection: Detectron2, EfficientDet, YOLO, and Faster R-CNN. The performance comparison of weed detection models is executed on the Open CV and Keras platform using python language.

Highlights

Object detection is one of the most active fields of research in computer vision, where it involves both object classification, classifying every object in the image, and object localization [1]
3.2 Results obtained for the YOLOv5 model After the end of the YOLOv5 training on 200 epochs, we evaluate its performance by the precision and error metrics on Tensorboard as follow: Fig. 15.The precision of the YOLOv5 model
The YOLOv5 training is faster than the other models, approximately 2 hours of training

Summary

Introduction

Object detection is one of the most active fields of research in computer vision, where it involves both object classification, classifying every object in the image, and object localization [1]. Agriculture is a field affected by these innovations to promote production and guarantee food security [2]. The focus is on the identification of weed from crop because of its great importance in precision farming, as weed act as a pest to crop and competes for space, nutrients, water, light and hinders the growth of crops in the field. The conventional way of eliminating weed is to spray herbicides or manual plucking [3]. The manual weed removal method is a tedious task, as it needs vast labor work. Usage of herbicides harms the health of living beings and the surrounding environment

Methods

Results

Conclusion