Spatio-temporal analysis for obstacle detection in agricultural videos

Campos, Yerania; Sossa, Humberto; Pajares, Gonzalo

doi:10.1016/j.asoc.2016.03.016

Cited by 32 publications

(8 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, their limited capacity to provide spectral and structural information is a deterrent to their usage in plant reconstruction. In addition, under outdoor conditions, the variable and uncontrolled illumination and the presence of shadows may represent a serious problem [ 10 ].…”

Section: Introductionmentioning

confidence: 99%

Influence of Wind Speed on RGB-D Images in Tree Plantations

Andújar

Dorado

Bengochea-Guevara

et al. 2017

Sensors

View full text Add to dashboard Cite

Weather conditions can affect sensors’ readings when sampling outdoors. Although sensors are usually set up covering a wide range of conditions, their operational range must be established. In recent years, depth cameras have been shown as a promising tool for plant phenotyping and other related uses. However, the use of these devices is still challenged by prevailing field conditions. Although the influence of lighting conditions on the performance of these cameras has already been established, the effect of wind is still unknown. This study establishes the associated errors when modeling some tree characteristics at different wind speeds. A system using a Kinect v2 sensor and a custom software was tested from null wind speed up to 10 m·s−1. Two tree species with contrasting architecture, poplars and plums, were used as model plants. The results showed different responses depending on tree species and wind speed. Estimations of Leaf Area (LA) and tree volume were generally more consistent at high wind speeds in plum trees. Poplars were particularly affected by wind speeds higher than 5 m·s−1. On the contrary, height measurements were more consistent for poplars than for plum trees. These results show that the use of depth cameras for tree characterization must take into consideration wind conditions in the field. In general, 5 m·s−1 (18 km·h−1) could be established as a conservative limit for good estimations.

show abstract

Section: Introductionmentioning

confidence: 99%

Influence of Wind Speed on RGB-D Images in Tree Plantations

Andújar

Dorado

Bengochea-Guevara

et al. 2017

Sensors

View full text Add to dashboard Cite

show abstract

“…In an effort to circumvent these factors researchers have enhanced standard semantic segmentation by embedding spatio-temporal information [29,9] into their architectures. In these cases, where spatio-temporal information was available, integrating this information improved performance [1,18]. Furthermore, Jarvers and Neumann [11] found that by incorporating sequential information, errors which occur in one frame could be recovered in subsequent frames.…”

Section: Related Workmentioning

confidence: 99%

Virtual Temporal Samples for Recurrent Neural Networks: applied to semantic segmentation in agriculture

Ahmadi¹,

Halstead²,

McCool³

2021

Preprint

View full text Add to dashboard Cite

This paper explores the potential for performing temporal semantic segmentation in the context of agricultural robotics without temporally labelled data. We achieve this by proposing to generate virtual temporal samples from labelled still images. This allows us, with no extra annotation effort, to generate virtually labelled temporal sequences. Normally, to train a recurrent neural network (RNN), labelled samples from a video (temporal) sequence are required which is laborious and has stymied work in this direction. By generating virtual temporal samples, we demonstrate that it is possible to train a lightweight RNN to perform semantic segmentation on two challenging agricultural datasets. Our results show that by training a temporal semantic segmenter using virtual samples we can increase the performance by an absolute amount of 4.6 and 4.9 on sweet pepper and sugar beet datasets, respectively. This indicates that our virtual data augmentation technique is able to accurately classify agricultural images temporally without the use of complicated synthetic data generation techniques nor with the overhead of labelling large amounts of temporal sequences.

show abstract

“…Spatial and temporal analyses were applied in video sequences in obstacle detection for safety purposes in [25]. The spatial analysis is based on the b* channel in the CIELAB color space where most objects can be distinguished from the main structures (plants and soil).…”

Section: Security: Obstacle Detectionmentioning

confidence: 99%

Machine-Vision Systems Selection for Agricultural Vehicles: A Guide

et al. 2016

Self Cite

View full text Add to dashboard Cite

Machine vision systems are becoming increasingly common onboard agricultural vehicles (autonomous and non-autonomous) for different tasks. This paper provides guidelines for selecting machine-vision systems for optimum performance, considering the adverse conditions on these outdoor environments with high variability on the illumination, irregular terrain conditions or different plant growth states, among others. In this regard, three main topics have been conveniently addressed for the best selection: (a) spectral bands (visible and infrared); (b) imaging sensors and optical systems (including intrinsic parameters) and (c) geometric visual system arrangement (considering extrinsic parameters and stereovision systems). A general overview, with detailed description and technical support, is provided for each topic with illustrative examples focused on specific applications in agriculture, although they could be applied in different contexts other than agricultural. A case study is provided as a result of research in the RHEA (Robot Fleets for Highly Effective Agriculture and Forestry Management) project for effective weed control in maize fields (wide-rows crops), funded by the European Union, where the machine vision system onboard the autonomous vehicles was the most important part of the full perception system, where machine vision was the most relevant. Details and results about crop row detection, weed patches identification, autonomous vehicle guidance and obstacle detection are provided together with a review of methods and approaches on these topics.

show abstract

Spatio-temporal analysis for obstacle detection in agricultural videos

Cited by 32 publications

References 24 publications

Influence of Wind Speed on RGB-D Images in Tree Plantations

Influence of Wind Speed on RGB-D Images in Tree Plantations

Virtual Temporal Samples for Recurrent Neural Networks: applied to semantic segmentation in agriculture

Machine-Vision Systems Selection for Agricultural Vehicles: A Guide

Contact Info

Product

Resources

About