Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging

Abstract Robotic harvesting systems face significant challenges in detecting sweet peppers due to the visual similarities between the fruit and its leaves in color and shape. Unlike fruits such as apples or tomatoes, where differences in shape and color facilitate straightforward identification thro...

Full description

Saved in:
Bibliographic Details
Main Authors: Zinat Tasneem, Koichi Oka, Naoya Tada
Format: Article
Language:English
Published: SpringerOpen 2025-08-01
Series:ROBOMECH Journal
Subjects:
Online Access:https://doi.org/10.1186/s40648-025-00317-2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849235026564087808
author Zinat Tasneem
Koichi Oka
Naoya Tada
author_facet Zinat Tasneem
Koichi Oka
Naoya Tada
author_sort Zinat Tasneem
collection DOAJ
description Abstract Robotic harvesting systems face significant challenges in detecting sweet peppers due to the visual similarities between the fruit and its leaves in color and shape. Unlike fruits such as apples or tomatoes, where differences in shape and color facilitate straightforward identification through visible imaging, sweet pepper detection requires more advanced techniques. This study introduces a distinctive method that leverages the temperature differential between sweet peppers and their leaves, primarily due to the hollow structure of the fruit. The hollow interior of sweet peppers, which contains air, causes the fruit’s temperature to change more slowly compared to the leaves, whose temperature fluctuates rapidly due to transpiration. This difference in thermal behavior under varying greenhouse conditions, including temperature, humidity, and illuminance, is exploited for detection. Two experimental phases were conducted to assess the effectiveness of this detection approach. The first experiment utilized a far-infrared camera, alongside temperature, humidity, and illuminance sensors, to capture and analyze thermal variations between the sweet peppers and their surrounding leaves. The results revealed a significant temperature differential, serving as a reliable indicator for fruit detection. In the second experiment, a depth camera was integrated to enhance the detection process, enabling the identification of sweet peppers within the robotic arm’s reachable range. This integration aimed to enhance the system’s ability to distinguish sweet peppers from leaves in three-dimensional space, improving detection precision for robotic harvesting. The experimental results demonstrate that the proposed thermal-based detection method shows significant promise for accurately identifying sweet peppers in greenhouse environments. This approach offers a valuable solution for robotic harvesting systems, particularly for crops where conventional visible imaging techniques are insufficient. However, the method’s applicability may be limited to hollow fruits like sweet peppers, as fruits with compact structures may not exhibit similar thermal characteristics.
format Article
id doaj-art-d2653ad27fcd427586fa23775211fa29
institution Kabale University
issn 2197-4225
language English
publishDate 2025-08-01
publisher SpringerOpen
record_format Article
series ROBOMECH Journal
spelling doaj-art-d2653ad27fcd427586fa23775211fa292025-08-20T04:02:55ZengSpringerOpenROBOMECH Journal2197-42252025-08-0112111910.1186/s40648-025-00317-2Sweet pepper detection in day and night greenhouse environments using thermal and depth imagingZinat Tasneem0Koichi Oka1Naoya Tada2Department of Intelligent Mechanical Systems Engineering, Kochi University of TechnologyDepartment of Intelligent Mechanical Systems Engineering, Kochi University of TechnologyDepartment of Intelligent Mechanical Systems Engineering, Kochi University of TechnologyAbstract Robotic harvesting systems face significant challenges in detecting sweet peppers due to the visual similarities between the fruit and its leaves in color and shape. Unlike fruits such as apples or tomatoes, where differences in shape and color facilitate straightforward identification through visible imaging, sweet pepper detection requires more advanced techniques. This study introduces a distinctive method that leverages the temperature differential between sweet peppers and their leaves, primarily due to the hollow structure of the fruit. The hollow interior of sweet peppers, which contains air, causes the fruit’s temperature to change more slowly compared to the leaves, whose temperature fluctuates rapidly due to transpiration. This difference in thermal behavior under varying greenhouse conditions, including temperature, humidity, and illuminance, is exploited for detection. Two experimental phases were conducted to assess the effectiveness of this detection approach. The first experiment utilized a far-infrared camera, alongside temperature, humidity, and illuminance sensors, to capture and analyze thermal variations between the sweet peppers and their surrounding leaves. The results revealed a significant temperature differential, serving as a reliable indicator for fruit detection. In the second experiment, a depth camera was integrated to enhance the detection process, enabling the identification of sweet peppers within the robotic arm’s reachable range. This integration aimed to enhance the system’s ability to distinguish sweet peppers from leaves in three-dimensional space, improving detection precision for robotic harvesting. The experimental results demonstrate that the proposed thermal-based detection method shows significant promise for accurately identifying sweet peppers in greenhouse environments. This approach offers a valuable solution for robotic harvesting systems, particularly for crops where conventional visible imaging techniques are insufficient. However, the method’s applicability may be limited to hollow fruits like sweet peppers, as fruits with compact structures may not exhibit similar thermal characteristics.https://doi.org/10.1186/s40648-025-00317-2Depth imageNighttime detectionRobotic harvestingSweet pepperTemperature differentialThermal imaging
spellingShingle Zinat Tasneem
Koichi Oka
Naoya Tada
Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
ROBOMECH Journal
Depth image
Nighttime detection
Robotic harvesting
Sweet pepper
Temperature differential
Thermal imaging
title Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
title_full Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
title_fullStr Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
title_full_unstemmed Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
title_short Sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
title_sort sweet pepper detection in day and night greenhouse environments using thermal and depth imaging
topic Depth image
Nighttime detection
Robotic harvesting
Sweet pepper
Temperature differential
Thermal imaging
url https://doi.org/10.1186/s40648-025-00317-2
work_keys_str_mv AT zinattasneem sweetpepperdetectionindayandnightgreenhouseenvironmentsusingthermalanddepthimaging
AT koichioka sweetpepperdetectionindayandnightgreenhouseenvironmentsusingthermalanddepthimaging
AT naoyatada sweetpepperdetectionindayandnightgreenhouseenvironmentsusingthermalanddepthimaging