Recognition of Wheat Spike from Field Based Phenotype Platform Using Multi-Sensor Fusion and Improved Maximum Entropy Segmentation Algorithms

Zhou, Chengquan; Liang, Dong; Yang, Xiaodong; Xu, Bo; Yang, Guijun

doi:10.3390/rs10020246

Cited by 55 publications

(38 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent developments in the application of the unmanned aerial vehicle (UAV) mounted with high definition cameras have increased the sample size tremendously [8][9][10]. Researchers have implemented many applications in plant height estimation [11][12][13], seedling counting [14][15][16], and crop growth estimation [17,18] using UAV images. Nevertheless, there are fewer applications of maize tassel detection using UAV images [19] which is challenging in natural environments due to light conditions, possible occlusions, and different maize genotypes.…”

Section: Introductionmentioning

confidence: 99%

Detection of Maize Tassels from UAV RGB Imagery with Faster R-CNN

et al. 2020

View full text Add to dashboard Cite

Maize tassels play a critical role in plant growth and yield. Extensive RGB images obtained using unmanned aerial vehicle (UAV) and the prevalence of deep learning provide a chance to improve the accuracy of detecting maize tassels. We used images from UAV, a mobile phone, and the Maize Tassel Counting dataset (MTC) to test the performance of faster region-based convolutional neural network (Faster R-CNN) with residual neural network (ResNet) and a visual geometry group neural network (VGGNet). The results showed that the ResNet, as the feature extraction network, was better than the VGGNet for detecting maize tassels from UAV images with 600 × 600 resolution. The prediction accuracy ranged from 87.94% to 94.99%. However, the prediction accuracy was less than 87.27% from the UAV images with 5280 × 2970 resolution. We modified the anchor size to [852, 1282, 2562] in the region proposal network according to the width and height of pixel distribution to improve detection accuracy up to 89.96%. The accuracy reached up to 95.95% for mobile phone images. Then, we compared our trained model with TasselNet without training their datasets. The average difference of tassel number was 1.4 between the calculations with 40 images for the two methods. In the future, we could further improve the performance of the models by enlarging datasets and calculating other tassel traits such as the length, width, diameter, perimeter, and the branch number of the maize tassels.

show abstract

Section: Introductionmentioning

confidence: 99%

Detection of Maize Tassels from UAV RGB Imagery with Faster R-CNN

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The use of RGB images may have limitations under certain field conditions, including the quality of the sky and light conditions, which can be overcome with sufficiently high spatial resolution, but which requires powerful computing capacities and makes its implementation more complex or less high throughput than expected. Other remote sensing approaches include the use of multispectral images [14], but the segmentation accuracy decreases as the canopy area observed within a single image increases, potentially due to the lower spatial resolution of these images and the reflectance angle dependence of multispectral data. Even LIDAR may be used [38] but its price and processing requirements may still be considered prohibitive and its size and weight makes it too cumbersome to be handheld or pole mounted for quick ground evaluation in field conditions.…”

Section: Discussionmentioning

confidence: 99%

“…In the manual in situ counting in the field, it was necessary to both view the canopy from different angles as well as physically move plants to acquire accurate field validation data, representing a major difference between the in situ counting and the single image-perspective remote sensing approach of the automatic thermal image ear counting technique presented here. In previous studies on ear recognition, no information regarding the correlation between in situ visual ear counting and automatic ear counting was provided [6][7][8][9][10][11][12][13][14], but it is nonetheless an important point to consider as the entire image acquisition and processing pipeline represents a sum of errors. Of course, the approach for visual counting assayed was in fact much faster than the traditional ear counting procedures, which implies for example counting the total number of ears in one-meter linear row length.…”

Section: Discussionmentioning

confidence: 99%

“…Depends of the hour of the day, 8:00 to 18:30 [10], 8:00 to 17:00 [14], 9:00 to 16:00 [9], 12:00 to 16:00 and sky (preferably diffuse light) conditions [12]. Possible algorithm errors -The algorithm presents errors when the air temperature is too low or high or the sky is too cloudy, or the conditions very windy, which may prevent differences between the canopy and the ear temperature appears.…”

Section: Stage Growthmentioning

confidence: 99%

“…Recently, a fusion of multispectral and RGB images have been developed for ear counting estimation [14].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Automatic Wheat Ear Counting Using Thermal Imagery

et al. 2019

View full text Add to dashboard Cite

Ear density is one of the most important agronomical yield components in wheat. Ear counting is time-consuming and tedious as it is most often conducted manually in field conditions. Moreover, different sampling techniques are often used resulting in a lack of standard protocol, which may eventually affect inter-comparability of results. Thermal sensors capture crop canopy features with more contrast than RGB sensors for image segmentation and classification tasks. An automatic thermal ear counting system is proposed to count the number of ears using zenithal/nadir thermal images acquired from a moderately high resolution handheld thermal camera. Three experimental sites under different growing conditions in Spain were used on a set of 24 varieties of durum wheat for this study. The automatic pipeline system developed uses contrast enhancement and filter techniques to segment image regions detected as ears. The approach is based on the temperature differential between the ears and the rest of the canopy, given that ears usually have higher temperatures due to their lower transpiration rates. Thermal images were acquired, together with RGB images and in situ (i.e., directly in the plot) visual ear counting from the same plot segment for validation purposes. The relationship between the thermal counting values and the in situ visual counting was fairly weak (R2 = 0.40), which highlights the difficulties in estimating ear density from one single image-perspective. However, the results show that the automatic thermal ear counting system performed quite well in counting the ears that do appear in the thermal images, exhibiting high correlations with the manual image-based counts from both thermal and RGB images in the sub-plot validation ring (R2 = 0.75–0.84). Automatic ear counting also exhibited high correlation with the manual counting from thermal images when considering the complete image (R2 = 0.80). The results also show a high correlation between the thermal and the RGB manual counting using the validation ring (R2 = 0.83). Methodological requirements and potential limitations of the technique are discussed.

show abstract

Biotic and abiotic factors governing dune response to storm events

Garzon

Ferreira

2021

Earth Surf Processes Landf

View full text Add to dashboard Cite

The alongshore response of dunes to storm events can be extremely variable and, consequently, their capacity to maintain their services, including the protection of hinterland communities. In this study, the role of biotic and abiotic factors determining the magnitude of dune retreat driven by a severe storm along a 60 km barrier island system was investigated. Data from high-resolution satellite imagery, digital terrain models, and wave propagation models were used in this assessment. The assessed abiotic factors included the backshore volume, dune height, downdrift inlet distance, and incident wave power. The evaluated biotic factor was the vegetation cover, characterized by a vegetation index retrieved from the multispectral imagery.The results revealed large alongshore variability on dune retreat, ranging from negligible impact to ca. 40 m of retreat. All combined factors allowed us to explain up to 70% of the dune retreat variability through a multi-regression analysis. Among all investigated factors, the major contributor controlling the magnitude of dune retreat was the backshore volume (more robust berms reduced the retreat) followed by the wave power (normal and longitudinal components). Moreover, the removal of local salient features in the dune line caused the straightening of the coastline, highly contributing to the development of dune retreat hotspots. The other evaluated factors had a smaller influence on reducing coastal retreat, including the vegetation, whose contribution to dune protection was around one order of magnitude lower than that provided by the backshore volume. The results highlight the importance of regional assessments to understand the causes behind the large alongshore variability of storm impacts at dunes. They also state the relatively low influence of the vegetation from this climatic region to enhance dune resistance to storms.

show abstract

Recognition of Wheat Spike from Field Based Phenotype Platform Using Multi-Sensor Fusion and Improved Maximum Entropy Segmentation Algorithms

Cited by 55 publications

References 55 publications

Detection of Maize Tassels from UAV RGB Imagery with Faster R-CNN

Detection of Maize Tassels from UAV RGB Imagery with Faster R-CNN

Automatic Wheat Ear Counting Using Thermal Imagery

Biotic and abiotic factors governing dune response to storm events

Contact Info

Product

Resources

About