Using a Camera System for the In-Situ Assessment of Cordon Dieback due to Grapevine Trunk Diseases

Background and Aims . Te assessment of grapevine trunk disease symptoms is a labour-intensive process that requires experience and is prone to bias. Methods that support the easy and accurate monitoring of trunk diseases will aid management decisions. Methods and Results . An algorithm was developed for the assessment of dieback symptoms due to trunk disease which is applied on a smartphone mounted on a vehicle driven through the vineyard. Vine images and corresponding expert ground truth assessments (of over 13,000 vines) were collected and correlated over two seasons in Shiraz vineyards in the Clare Valley, Barossa, and McLaren Vale, South Australia. Tis dataset was used to train and verify YOLOv5 models to estimate the percentage dieback of cordons due to trunk diseases. Te performance of the models was evaluated on the metrics of highest confdence, highest dieback score, and average dieback score across multiple detections. Eighty-four percent of vines in a test set derived from an unseen vineyard were assigned a score by the model within 10% of the score given by experts in the vineyard. Conclusions . Te computer vision algorithms were implemented within the phone, allowing real-time assessment and row-level mapping with nothing more than a high-end mobile phone. Signifcance of the Study . Te algorithms form the basis of a system that will allow growers to scan their vineyards easily and regularly to monitor dieback due to grapevine trunk disease and will facilitate corrective interventions.


Introduction
Grapevine trunk diseases (GTDs), such as Eutypa and Botryosphaeria dieback, are a pervasive and growing issue across the Australian wine industry that gradually reduces vineyard performance.Other trunk diseases, such as esca, Petri disease, Phomopsis dieback, and black foot disease, cause signifcant issues in other countries but have little impact in Australia [1].Eutypa dieback causes leaves to become distorted and yellow, shoots to stunt, and cordons to dieback.Botryosphaeria dieback has no distinct foliar symptoms but causes similar cordon dieback.GTDs are detected by the visual assessment of experts, and the control treatments for GTDs can be labour-intensive and most efective when administered preventively, early in the life of the vineyard [1][2][3][4].Regular vineyard surveys are not feasible for many growers due to the labour resources required.
Methods for estimating GTD dieback from aerial imagery are well-established but are limited by ground vegetation [5].Recent work by Ouyang et al. [6] used 3D point clouds collected using an unmanned aerial vehicle to detect GTD with an accuracy of 87.4%.
Deep learning techniques are part of a rapidly growing area of machine learning research that is especially efective for image analysis such as the detection of GTD.Deep learning methods typically result in higher classifcation accuracy and faster testing times than using traditional machine learning methods, but most critically for this research, they eliminate the need for hand-crafted features [7].Tese advantages over traditional machine learning have caused deep learning image analysis to be used in a wide variety of agricultural applications, including disease identifcation [8][9][10][11].Researchers applied combinations of different networks, both existing and custom architectures, on datasets that they had collected and augmented themselves [7-9, 11, 12].
Mohanty et al. [12] previously achieved an overall accuracy of 99.35% when detecting crop-disease pairs from images of leaves using DL techniques, but there are key diferences in the scope of research programs.Tey identifed 26 diferent diseases in 14 crop species, but the images used were of a single leaf, taken in a controlled environment against a consistent background.Our research aims to detect the presence of a single disease in real time from images taken in the feld, which introduces a number of complications.Te in-feld images introduce uncontrolled backgrounds and conditions, which can reduce the accuracy of the detection.
For in-feld images, there has been a wide variety of work in object detection for agriculture, most notably fruit detection.Kuznetsova et al. [13] applied the YOLOv5 algorithms for apple detection with a false positive rate of 3.5% and a false negative rate of 2.8%.For strawberry detection, Chen et al. [14] achieved a false positive rate of 5.7 to 15.4% and a false negative rate between 4.6% and 18.1% on mature fruit.Wang et al. [15] studied various attributes of fruit detection using YOLOv5 and recommended that for singleclass object detection, a minimum of 2500 objects should be labelled and used in training.
Beyond object detection, the classifcation of severity or other fruit attributes has also been studied.In addition to their mature strawberry detection, Chen et al. [14] investigated fower and immature fruit detection, with limited success.Wang et al. [16] adapted a VGG-16 classifcation model for estimating apple fower distributions, focussing on the maturity stage rather than the frequency of each class.Tey showed it to be more accurate and slightly faster than YOLOv5 when running on a personal computer.
Te aim of this research was to develop an automated edge computing system that would allow growers to quantify the severity of cordon dieback caused by GTDs at a temporal (every season) and spatial (whole vineyard) scale.Te system had to use a standard camera mounted on a vineyard vehicle and intelligent algorithms to monitor and map trunk disease and be implemented in such a way that it could be accessed by nontechnical users.
Te aim can be split into two components: (1) Algorithms for cordon dieback assessment (2) System for data collection, processing, and display Tis paper presents the frst component of the research and evaluates its performance.Te algorithm for cordon dieback assessment will be a machine-learning-based image processing algorithm trained using vineyard images collected on a standard camera and will be assessed on the similarity of the algorithm's results to expert assessment on unseen vines.Te scope of this research is limited to vines with bilateral cordons with spurs, trained on a single wire, due to these being more common than quadrilateral cordons in an Australian context.

Materials and Methods
2.1.Data.Te data used to evaluate the dieback assessment networks were collected in October of 2020 and 2021 in eleven vineyards (cv.Shiraz, Vitis vinifera L.) in the McLaren Vale and the Clare and Barossa Valleys, South Australia.Images of the vines were collected using a mobile phone app developed for the purpose and operating on a pair of Samsung Galaxy S21+ phones (model SM-G996B) running Android 11.Tese phones were mounted on a trailer approximately 300 mm from the ground and the middle of the interrow, with the image sensor facing the vines and the phone orientated so the cordon wire was near the centre of the image.See Figure 1 for the experimental setup.Te trailer was driven throughout the vine rows at a speed of approximately 7-9 km/h while imagery was captured and processed by the phone.Images were captured by each phone at a rate of at least 5 frames per second and a resolution of 1280 × 720 pixels.When combined with the wide feld of view lens in the phone, this enabled the majority of each vine to be captured, with the trunk at the centre of the image.Further analysis of the achievable framerate is given in Section 4.3.
Te proportion of cordon dieback on each vine was also visually assessed by two experts in the vineyard, and the score was recorded for each of the assessed vines [4,17].Cordon dieback in these vineyards is predominantly caused by GTDs, as evidenced by the presence of Eutypa dieback foliar symptoms, but it should be acknowledged that other factors such as nematodes, viruses, and other vineyard management practices may have contributed to the cordon dieback [18].Each cordon was assigned a score in the range of (0, 50) in increments of 5, representing the percentage of dieback on the cordon as a total of the vine.Class 0 represents a complete and healthy canopy, and class 50 represents a cordon with no shoots or leaves.Te assessment of dieback can vary between experts, and there is a particular difculty in diferentiating between the lower classes of 0, 5, and 10.Tese scores were matched with the images of the vines, and the images were labelled with bounding boxes around the trunk and around each cordon with the dieback score.During the growing season for the 2021 vintage, 12,642 bilateral cordon vines were scored and imaged, with  Data augmentation techniques were used to increase the number of training examples, so that the network would be more robust to changes in orientation and variable environmental conditions.Each training image was fipped horizontally with a probability of 85%, which would simulate driving the vehicle carrying the camera in each direction along the row of vines, capturing images of both sides of the vine.A Gaussian blur was applied to the training images to increase the number of training instances and to increase the robustness of the algorithm to lower-quality images which may occur when capturing images from a moving platform.Te weather conditions greatly afected the brightness of the grapevine images, so each of the images had its brightness both increased and decreased using a gamma correction function to simulate a range of weather conditions.Gamma correction applies a mathematical function to each pixel that either lightens or darkens the image overall, depending on the parameters used.Te augmentations applied increased Figure 1: Two pairs of smartphones mounted facing the opposite rows for continuous data collection.Tis prototype setup also includes two action cameras and two vertically orientated phones; note that only images collected from the horizontally orientated phones were included here.Other than a companion phone to control these phones from the driver's seat, no additional infrastructure was needed.See Figure 2 for an example image.the number of training images from 2084 images to 13076 images.Te validation and test set images were not augmented in any way (Table 1).
Experiments were carried out to evaluate the suitability of the proposed algorithm by varying the hyperparameters and the data used for training each model (Supplementary Table 1 to Supplementary Table 6).Te Ultralytics YOLOv5 version 6.1 Python library was used to implement the algorithm [19].Training was carried out on a personal computer with 16 Intel ® Core ™ i9-9900KF CPUs using Python 3.7.3 and Ubuntu 18.04.6LTS.Te data used to train a deep learning image processing network is crucial and one of the defning factors in the results.
All models were evaluated on an unseen test set consisting of all the assessed vines in one block.Tis was to ensure that there was no overlap between the training and test data and that the results of each experiment could be directly comparable.Te primary variables that were investigated were the data used to train the network and the training hyperparameters.All experiments were trained to completion, with completion being defned as the trend of the accuracy on the validation set across training epochs appearing to stabilise, with training lasting at least 300 epochs.
Te most accurate model was evaluated not only on the unseen test set (Block 4) but also on a much larger set of images from the remaining blocks, again ensuring that these images were not included in either the training or validation sets.

Algorithm Evaluation Metrics.
Te success of the dieback assessment algorithm was measured using the following criteria: (i) Percentage of trunks detected (ii) Percentage of cordons detected (iii) Percentage of cordons with dieback scores identifed correctly (class accuracy) (iv) Percentage of cordons with dieback scores identifed within 5% of correct score (variation accuracy ±5%) (v) Percentage of cordons with dieback scores identifed with 10% of correct score (variation accuracy ±10%) Te percentage of trunks detected should be as high as possible, as the system used to analyse the images relies on the detection of a trunk or half cordon to denote the results of the dieback assessment algorithm.By detecting the trunk and using images only where the trunk appeared close to the centre of the image, double-counting of successive half cordons was avoided.Te algorithm must be able to detect the grape vine cordons in order to identify the extent of dieback, so the successful detection of cordons must occur for the algorithm to be efective.Te assessment of the extent of dieback is subjective and can vary between experts.Terefore, the identifcation of the dieback score for each cordon will be assessed on an exact match to the in-feld scoring as well as with a margin of 5% or 10% error.

System Overview.
To manage, control, and observe the scanning process with ease, a smartphone-based twoapplication system was designed with a "controller" and a "scanner" application (Figures 4 and 5).Te system only needs to connect to external devices on two occasions: for the initial fast localisation of the GNSS system or when downloading the map data for display on a computer.Te system is able to process the images and automatically generate a map of the GTD in real time using only the "scanner" phone, the results of which are displayed on the "controller" phone or a computer.Further details of the system are outside the scope of this paper and available on request.

Dieback Assessment Algorithm.
Following the experiments used for training the dieback assessment algorithms, model 6 gave the best overall performance (Table 2).Te trunk class was excluded from the confusion matrix (Figure 6) for the best performing model (model six) as all trunks were correctly detected in the test set.Missing cordons, which are cordons that were labelled, but not detected by the algorithm, were designated a separate class ("M") in the confusion matrix.Model 6 was applied to images collected in the same blocks used for training.Even though these vines and images were not seen by the model during training or validation, excellent correlation with ground truth is seen, with over 99% of vines having an estimated GTD dieback severity within 10% of the manual ground truth (Table 2 and Figure 6(b)).
When the most successful model (model 6) was applied to the unseen test set; that is, with vines from a block completely unseen in the training or validation, the shape of the distribution is well matched against ground truth data (Figure 7(a)).Similar patterns were seen for the blocks used as part of the validation (within the training process) (Supplementary Figure 1).Examples of detections in images are shown in Figures 8 and 9.

Evaluation of the Selected Model across Eleven Test Sites.
Data from the eleven sites used for training and validating the algorithm were processed with model 6 using the smartphone-with an additional block in the Barossa Valley also mapped (Block 1).Histograms were used to display the distribution of GTD severity across the block (Supplementary Figure 1).Te vines and severity of GTD were georeferenced and plotted on aerial images (Figures 10-12).
In Block 1 (Figure 10), the mapped data displayed a high degree of average severity uniformly distributed across all of the surveyed vines.Whilst there are pockets of higher-  In Block 5 (Figure 11), a high concentration of vines exhibiting severe symptoms were located at the northern end of the rows.Grapevine trunk disease does not normally follow a spatial pattern-so the grouping of the afected vines in one section of the vineyard was surprising.On further investigation, it was identifed that the northern end of the block had reduced vigour as it is prone to frost, and a frost event had occurred several weeks before the assessment.Regardless of the cause, this gives growers an indicator that this is an area where the vines are performing poorly.A further manual inspection would often be made of the worst-afected areas to confrm the cause of an unusually concentrated area of increased dieback.
Te mapping of Block 8 (Figure 12) exhibits less severe symptom severity.Te high-symptom severity vines are clustered into small groups and distributed across the eastern portions of the block.
Vine symptom severity was usually normally distributed across the respective block, with a skew towards lower levels of severity (Figure 13).Te results across blocks were typically clustered to a 10-20% range with some outliers.Blocks 3 and 5 exhibit results with a wider spread, with lower peaks, and a fatter distribution.In Block 5, this was a cause of the severe concentration of vine symptom severity in a small section of the block (see Figure 11).

Application Performance and Optimisation.
Te target framerate (5 FPS) was achieved consistently as a result of optimisation of the phone application.Images were captured at 1280 × 720 pixels and processed at 640 × 360 pixels using model 6.Te two test phones used (128 GB and 256 GB models of the SM-G996B Samsung Galaxy S21+ 5G) were both able to maintain a throughput of at least 5 FPS, shown as the ability to process individual images consistently in less than 200 ms over 110 minutes (Figure 14).Te increase in processing time observed in the 128 GB model at the 60 minute mark is likely due to processor throttling as the phone heated up over time; however, the 200 ms threshold was not exceeded.

Trunk and Cordon Detection.
Trunk detection was high across all the experiments, with at least 97% of trunks being detected in each experiment and trunk detection as high as 100% in two of the trained models.Trunk detection was consistently high because of the number of instances in the training data and the appearance of the trunks.Te trunks are visually distinct from the cordons, most notably due to their orientation.For every grapevine, there is a single trunk and two bilateral cordons with spurs; quadrilateralcordon vines were considered out of scope for this research due to their distinctly diferent appearance.Given that the cordons are broken down into 10 classes based on the extent of dieback, the number of instances of trunks is much higher than any other class.Deep learning object detection algorithms require many examples to accurately detect objects in images; therefore, the high number of training instances for trunks ensures that the trunk detection was successful.
Te algorithm must detect the grapevine cordons in order to classify them based on the extent of dieback, which makes the percentage of cordons detected critical to the overall performance of the algorithm.Two cordons were not detected in the test set.In the frst example of a missing cordon, the cordon was not detected as there was a tree in the background with the foliage extending above and below the cordon, so the cordon was not distinguished from the background (Figure 15).For the second missed cordon detection, the photo is blurred and the leaves are pale in the image, but the cordon is not unrecognisable to a human observer (Figure 16).Tere are other considerations for the algorithm in this example.First, the right cordon was detected with a confdence score of 0.41, low compared to the majority of cordon confdence scores, which suggests that the light conditions and the blur (more  8 Australian Journal of Grape and Wine Research signifcant on the left cordon) were a major factor in any detections of the left cordon falling under the confdence threshold of 0.25.Te bounding box for the trunk is also much wider than the typical trunk bounding boxes as the trunk itself is slanted (Figure 16).Te cordons do not originate from the centre of the trunk bounding box, as is the norm, and combined with the thinness of the left cordon, this creates additional difculties for the detection of this cordon.
In terms of pure object detection accuracy using YOLOv5, Kuznetsova et al. [13] obtained an accuracy of 97.1% in counting apples in general images.It is not surprising that the trunk detection results in this work are slightly higher in accuracy given the size and uniqueness of the shape compared with apples.

GTD Dieback Detection.
Detecting trunks and cordons in vineyard images allows the algorithm to fulfll the aim of detecting the extent of dieback.Model 4 had the highest class accuracy, with 27% of cordons classifed by the algorithm matching the labels given in the vineyard (Table 2).As previously stated (see Section 3.1), the dieback scoring is subjective and can vary between different experts, and there is particular difculty in differentiating between classes 0, 5, and 10.When the variation accuracy within 5% and 10% was considered, model 6 had the highest ±10% variation accuracy (84%), as well as a higher ±5% variation accuracy and more cordons detected than model 4. Models 4 and 6 had slightly diferent training hyperparameters, but the main diference between these models was the training data      [15].However, the test set images were taken on a single day using one model of phone.Terefore, the efects of changing illumination due to weather conditions and changing the photo quality due to the camera used were not directly assessed.Te ±10% variation accuracy is quite reasonable given that some classes included as little as 50 training images (prior to augmentation), and Wang et al. [15] recommended 2500 training images for a single class.Te detection algorithm is more likely to overestimate the extent of the dieback rather than underestimate it, with 19 cordons underestimated by at least 15% and 28 cordons overestimated by at least 15%, the manual scoring factors in all the shoots extending from each cordon, including when a shoot extends over the adjacent cordon, although this is not common.Te detection algorithm estimates the extent of dieback based on the volume of leaves around the cordon, as the training images were labelled with a bounding box around the cordon, and the volume of leaves in the bounding box is largely consistent with the amount of dieback.If a shoot extends to another cordon, the algorithm will estimate the extent of dieback incorrectly (see Figure 2).Te right cordon in this example was given a manual score of 50, but the detection algorithm assigned a score of 30 due to the shoots from the adjacent cordon extending into the bounding box.Te accuracy of 84% on an individual vine level compares well with that of Ouyang et al. [6], who achieved 87% accuracy on an aggregated row level.Te number of classes of severity used by Ouyang et al. [6] was slightly smaller, which would also lead to improved results.
When the frequencies of each class in the manual scoring and detections were compared between a set of vines from a vineyard that was not used in training (Figure 7  Tis would not refect the manual scoring system as closely but may align more with the needs of the growers.Te presence of growth in the area is more important than exactly which vine it extends from.Alternatively, semantic segmentation could be used to identify exactly which shoots extend from each cordon to align with the manual scoring more closely.Semantic segmentation would not appear to currently be a feasible technique for real-time processing on a mobile phone due to the need to classify each pixel rather than identify three bounding boxes.Te training data would    also need to be labelled by experts in manually scoring dieback as correctly allocating each shoot to the correct cordon is very important.

Evaluation of the Selected Model across Eleven Test Sites.
By comparing the results across all the test sites, signifcant variation in the spatial pattern, incidence, and severity of GTD dieback symptoms was observed.Tis may be due to diferent ages of vines, diferent GTDs involved, or local climatic conditions.Te incidence and severity of GTDs increases with age of vines [3,4,20].Te distribution of pathogens that cause Eutypa and Botryosphaeria dieback varies between Australian regions [21,22] which may also explain some of the variability observed between regions in the current study.Rainfall is required for infection, and certain temperature and humidity conditions favour the diferent causal pathogen species [23,24].Nonetheless, this work lays the foundation for the analysis of data over multiple seasons or before and after remediation activities to monitor changes in GTD symptoms.Very few blocks have a low average severity (Supplementary Figure 1), partially due to grapevines being a natural system and not growing uniformly despite the best endeavours of growers.It also highlights the potential for growers to tend to underestimate the severity throughout their blocks, as once the canopy is more fully grown, shoots will tend to spread out and disguise diseased sections of the cordon.

Evaluation of Smartphone Application.
After running for many hours in-feld conditions, the smartphone "scanner" application was able to successfully collect, process, and georeference all the images across the eleven test sites.Despite the hard requirement of 5 FPS processing, the phone was able to sustain this performance consistently in tests lasting more than an hour.Compared with the aerial method of Ouyang et al. [6], the ability to undertake the survey using only a mobile phone mounted on a vehicle is somewhat simpler yet of comparable accuracy, giving greater opportunity for industry adoption.

Conclusions
Tis paper presents and evaluates an algorithm to detect and map grapevine trunk disease dieback using only a smartphone.Te YOLOv5-based algorithm was successfully applied in a smartphone app to collect and process data from more than 13,000 vines in the McLaren Vale, Clare Valley, and Barossa Valley regions of South Australia across two growing seasons and ten vineyards.
Te algorithm was efective, as it was able to classify 99% of cordons within 10% of expert visual dieback assessment on unseen vines from the same blocks as used in the training and validating the model.When tested on vines from a diferent block, again unseen by the model, a classifcation accuracy of 84% was achieved and 99.5% of cordons were detected.
Furthermore, the algorithm reliably operated at a frame rate of 5 FPS on a commercially available smartphone, including capturing, processing, and mapping the data with GNSS.
Further research into the robustness of the algorithm under diferent weather conditions and image quality is recommended to ensure that the system remains efective for many models of phone used and that the system is not reliant on good weather conditions.A variation of the algorithm that can be used in vineyards with diferent training systems (e.g., multiple cordons) would also be a recommended area of further research.A reliance on existing deep learning algorithms mean the GTD level had to be discretised; further work could examine methods for providing a continuous numerical output.
Being able to transform a deep learning model trained on a server to run in real-time on a smartphone has provided a powerful tool for growers to attach to a vehicle and obtain maps of GTD dieback symptoms.Tis opens the potential for rapid assessment of GTD more widely across the industry on bilateral cordon-trained vines.It also highlights the potential for deep learning models to be trained to detect visual symptoms of other diseases and to be applied in the feld with just a smartphone.

2
Australian Journal of Grape and Wine Research good detection results.Te total dataset of all scored vine image across each vineyard resulted in an unequal number of instances between cordon classes (Figure 3(a)).Tis class imbalance can result in the assessment network overftting to certain classes, artifcially increasing the probability of assessing certain classes.Tis is particularly detrimental to the classes with very few training examples, such as classes 45 and 50 (Figure 3(a)).A subset of the total dataset with a balanced class distribution was created and used as the balanced training dataset for the network (Figure 3(b)).Te number of training examples was greatly reduced when a balanced training dataset was created, as the number of training examples in each class was reduced to approximately the number of instances in the smallest class (class 50), and the majority of training examples consisted of classes 5, 10, and 15.Experimentation was used to explore the efects of various combinations of training sets across years.

Figure 2 :
Figure 2: Detection from the unseen test set.Te dieback on the right cordon is underestimated as the shoots from the adjacent cordon extend into the bounding box.

Figure 3 :
Figure 3: (a) Te distribution of vine cordon GTD labels for imagery collected in the 2022 vintage.(b) Te more balanced distribution of vine cordon GTD labels used for training the models, using data collected from both 2021 and 2022 vintages.Zero indicates no GTD symptoms observed, and 50 indicates complete dieback for that cordon.Te total number of cordons is approximately twice the number of vines as indicated by the number of trunks identifed.

Figure 5 :
Figure 5: System overview diagram.Te internal phone GNSS receiver is used to geotag each image.As such, the positional accuracy is dependent on having mobile phone reception on initial setup.

Figure 6 :Figure 7 :
Figure 6: Confusion matrices for cordon class detections for model 6 (as described in Table2).(a) Images for the unseen test set were collected from vineyards where other images from those vineyards were used in training.(b) Images for the unseen vines comprised of vines from Block 4 which had not been used for any model training or validation the yellow of-diagonal terms have been highlighted as a ± 10 threshold has been applied in the evaluation of results in this paper.Te general diagonal form of the results is evident, with some outliers (discussed below).

Figure 8 :
Figure 8: Detections from model 6.Te labelled classes were 15 on the left and 15 on the right.Te detected classes were 15 for the left cordon and 10 for the right, as the detected classes are given frst in the bounding box labels.Te decimal in the bounding box is the confdence score, an expression of how sure the algorithm is of what is detected.

Figure 9 :
Figure 9: Detections from model 6.Te labelled classes were 25 on the left and 30 on the right.Te detected classes were 20 for the left cordon and 30 for the right, as the detected classes are given frst in the bounding box labels.Te decimal in the bounding box is the confdence score, an expression of how sure the algorithm is of what is detected.

Figure 10 :
Figure 10: Block 1 vine symptom mapping-each point represents the overall vine assessment combining both cordons.

Figure 11 :
Figure 11: Block 5 vine symptom mapping-each point represents the overall vine assessment combining both cordons.

10
Australian Journal of Grape and Wine Research used.Model 6 used the V2021 and V2022 augmented data, resulting in many more training examples which improved the performance similarly to the cordon detection.Te augmented data adjusted the exposure and blur of the training images, which theoretically would make the algorithm more robust to changing light conditions and changes in photo quality, as shown by Wang et al.
(a)) and a larger test set consisting of the unseen vines in vineyards that were used in training the algorithm (Figure 7(b)), the algorithm performed better on the unseen vines in vineyards used in training.Te images of the unseen vineyard are often overexposed, although some of the unseen vines in the training blocks are overexposed as well, these are a higher proportion in the unseen vineyard.Te images in the unseen vineyard are blurred in addition to the light conditions that would cause more difculties in accurately estimating the dieback.Tere are two possible courses of action to potentially improve the results.Te training images could be given new scores based on the volume of leaves around each cordon.

Figure 12 :
Figure 12: Block 8 vine symptom mapping-each point represents the overall vine assessment combining both cordons.

Figure 13 :Figure 14 :
Figure 13: Cordon dieback severity distribution across blocks, relative to 25% of total number of vines scanned in that block.Subfgures (a-k) correspond to Blocks 1 to 11, respectively.

Figure 15 :
Figure 15: A missing cordon detection from the test set.

Figure 16 :
Figure 16: Another missing cordon detection from the test set.

Table 1 :
Number of images in each data subset.
-Change scanner settings (e.g.custom names, configured facing, etc.) Te percentage of detected cordons rose with the number of training examples and the increase in the left-right fip during training.Te increase of the left-right fip hyperparameter also efectively

Table 2 :
Results of training GTD detection algorithm when applied to a completely unseen test set (Block 4).
creasing the training examples diminished as more training examples were used, but the network achieved the correct detection of 99% of cordons in the unseen test set which underpins the rest of the analysis.