Abstract

The detection of cherry tomatoes in greenhouse scene is of great significance for robotic harvesting. This paper states a method based on deep learning for cherry tomatoes detection to reduce the influence of illumination, growth difference, and occlusion. In view of such greenhouse operating environment and accuracy of deep learning, Single Shot multi-box Detector (SSD) was selected because of its excellent anti-interference ability and self-taught from datasets. The first step is to build datasets containing various conditions in greenhouse. According to the characteristics of cherry tomatoes, the image samples with illumination change, images rotation and noise enhancement were used to expand the datasets. Then training datasets were used to train and construct network model. To study the effect of base network and the input size of networks, one contrast experiment was designed on different base networks of VGG16, MobileNet, Inception V2 networks, and the other contrast experiment was conducted on changing the network input image size of 300 pixels by 300 pixels, 512 pixels by 512 pixels. Through the analysis of the experimental results, it is found that the Inception V2 network is the best base network with the average precision of 98.85% in greenhouse environment. Compared with other detection methods, this method shows substantial improvement in cherry tomatoes detection.

Highlights

  • China is the world’s largest tomato production and consumption country

  • Considering different lighting conditions and various tomato growth states, an improved model for cherry tomatoes detection based on the Single Shot multi-box Detector (SSD) proposed by Liu et al [16] is presented in this paper

  • The separated cherry tomatoes mean that fruit are independent and without any obstruction

Read more

Summary

Introduction

China is the world’s largest tomato production and consumption country. The area of tomato plantation is about 1.0538 million hm in an average year, with a total production of 54.133 million Mg, which accounts for 21% of the world’s total production [1]. More and more research is focused on using machine learning or deep learning to solve the problem of fruit recognition.Wang et al [10] used a color-based K-means clustering algorithm to segment litchi images with an accuracy rate of 97.5%. Fu et al [14] stated a deep learning model based on LeNet convolutional neural network for multi-cluster kiwi fruit image detection, and the accuracy rate achieved 89.29%. Considering different lighting conditions and various tomato growth states, an improved model for cherry tomatoes detection based on the SSD proposed by Liu et al [16] is presented in this paper. The model is improved by using end-to-end training method to adaptively learn the characteristics of cherry tomatoes in different situations, which achieves fruit recognition and localization in unstructured harvesting environment

Image Acquisition
Sample Data Set
Classical SSD Deep Learning Model
Improved SSD Deep Learning Model
Overview of Detection Algorithm
Experiment Design
Experiment Parameters
Evaluation Standard
Experiment Results Analysis
Methods
Conclusions and Future Work
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call