A Bayesian Framework for the Assessment of Vision-based Weed and Fruit Detection and Classification Algorithms

This paper proposes new metrics and a performance-assessment framework for vision-based weed and fruit detection and classification algorithms. In order to compare algorithms, and make a decision on which one to use fora particular application, it is necessary to take into account that the performance obtained in a series of tests is subject to uncertainty. Such characterisation of uncertainty seems not to be captured by the performance metrics currently reported in the literature. Therefore, we pose the problem as a general problem of scientific inference, which arises out of incomplete information, and propose as a metric of performance the(posterior) predictive probabilities that the algorithms will provide a correct outcome for target and background detection. We detail the framework through which these predicted probabilities can be obtained, which is Bayesian in nature. As an illustration example, we apply the framework to the assessment of performance of four algorithms that could potentially be used in the detection of capsicums (peppers).

Application of commercially available, low-cost, miniaturised NIR spectrometers to the assessment of the sugar content of intact fruit

Recent decreases in costs, and improvements in performance, of silicon array detectors open a range of potential applications of relevance to plant physiologists, associated with spectral analysis in the visible and short-wave near infra-red (far-red) spectrum. The performance characteristics of three commercially available ‘miniature’ spectrometers based on silicon array detectors operating in the 650–1050-nm spectral region (MMS1 from Zeiss, S2000 from Ocean Optics, and FICS from Oriel, operated with a Larry detector) were compared with respect to the application of non-invasive prediction of sugar content of fruit using near infra-red spectroscopy (NIRS). The FICS–Larry gave the best wavelength resolution; however, the narrow slit and small pixel size of the charge-coupled device detector resulted in a very low sensitivity, and this instrumentation was not considered further. Wavelength resolution was poor with the MMS1 relative to the S2000 (e.g. full width at half maximum of the 912 nm Hg peak, 13 and 2 nm for the MMS1 and S2000, respectively), but the large pixel height of the array used in the MMS1 gave it sensitivity comparable to the S2000. The signal-to-signal standard error ratio of spectra was greater by an order of magnitude with the MMS1, relative to the S2000, at both near saturation and low light levels. Calibrations were developed using reflectance spectra of filter paper soaked in range of concentrations (0–20% w/v) of sucrose, using a modified partial least squares procedure. Calibrations developed with the MMS1 were superior to those developed using the S2000 (e.g. coefficient of correlation of 0.90 and 0.62, and standard error of cross-validation of 1.9 and 5.4%, respectively), indicating the importance of high signal to noise ratio over wavelength resolution to calibration accuracy. The design of a bench top assembly using the MMS1 for the non-invasive assessment of mesocarp sugar content of (intact) melon fruit is reported in terms of light source and angle between detector and light source, and optimisation of math treatment (derivative condition and smoothing function).

Non-invasive assessment of pineapple and mango fruit quality using near infra-red spectroscopy

The potential of near infra-red (NIR) spectroscopy for non-invasive measurement of fruit quality of pineapple (Ananas comosus var. Smooth Cayenne) and mango (Magnifera indica var. Kensington) fruit was assessed. A remote reflectance fibre optic probe, placed in contact with the fruit skin surface in a light-proof box, was used to deliver monochromatic light to the fruit, and to collect NIR reflectance spectra (760–2500 nm). The probe illuminated and collected reflected radiation from an area of about 16 cm2. The NIR spectral attributes were correlated with pineapple juice Brix and with mango flesh dry matter (DM) measured from fruit flesh directly underlying the scanned area. The highest correlations for both fruit were found using the second derivative of the spectra (d2 log 1/R) and an additive calibration equation. Multiple linear regression (MLR) on pineapple fruit spectra (n = 85) gave a calibration equation using d2 log 1/R at wavelengths of 866, 760, 1232 and 832 nm with a multiple coefficient of determination (R2) of 0.75, and a standard error of calibration (SEC) of 1.21 °Brix. Modified partial least squares (MPLS) regression analysis yielded a calibration equation with R2 = 0.91, SEC = 0.69, and a standard error of cross validation (SECV) of 1.09 oBrix. For mango, MLR gave a calibration equation using d2 log 1/R at 904, 872, 1660 and 1516 nm with R2 = 0.90, and SEC = 0.85% DM and a bias of 0.39. Using MPLS analysis, a calibration equation with R2 = 0.98, SEC = 0.54 and SECV = 1.19 was obtained. We conclude that NIR technology offers the potential to assess fruit sweetness in intact whole pineapple and DM in mango fruit, respectively, to within 1° Brix and 1% DM, and could be used for the grading of fruit in fruit packing sheds.

Activating Mango Fruit Defence to Anthracnose Disease

The fungus causing anthracnose disease in mango, Colletotrichum gloeosporioides, (C g.), infects immature fruit early in the season, then enters a long latent phase. After harvest, when fruit start to ripen, the latency breaks and the fungus ramifies through the peel and pulp tissues causing black disease lesions. The breaking of pathogen latency in ripening mango fruit has been correlated with decreasing concentrations of the endogenous antifungal resorcinol compounds (Droby et al., 1986). The level of these antifungal resorcinols vary among mango cultivars (Droby et a1 , 1986). Controlling diseases by managing natural resistance of fruit to fungal attack could minimize the use of pesticides, which have become of major public concern on health and environmental grounds. The plant resistance activator benzo(l,2,3)thiadiazole-7-carbothioic acid S-methyl ester (trade name Bion®) has been widely reported as an effective inducer of systemic resistance. For example, Bion® was reported to induce pathogenesis-related proteins (PR proteins) and stimulate plant defence in peas (Dann and Deverall, 2000) and roses (Suo and Leung, 2001). However, until now, there is no information about the role of Bion® in activation of mango (cv. Kensington Pride) fruit resistance to anthracnose disease. The aim of this research is to determine the effect of resistance activators on defence responses of mango fruit to anthracnose disease.

Effect of nitrogen on the skin colour and other quality attributes of ripe ‘Kensington Pride’ mango (Mangifera indica L.) fruit

Near-ripe ‘Kensington Pride’ mango (Mangifera indica L.) fruit with green skin colour generally return lower wholesale and retail prices. Pre-harvest management, especially nitrogen (N) nutrition, appears to be a major causal factor. To obtain an understanding of the extent of the problem in the Burdekin district (dry tropics; the major production area in Australia), green mature ‘Kensington Pride’ mango fruit were harvested from ten orchards and ripened at 20 ± 0.5 O C. Of these orchards, 70% produced fruit with more than 25% of the skin surface area green when ripe. The following year, the effect of N application on skin colour and other quality attributes was investigated on three orchards, one with a high green (HG) skin problem and two with a low green (LG) skin problem. N was applied at pre-flowering and at panicle emergence at the rate of 0,75,150,300 g per tree (soil applied) or 50 g per tree as foliar N for the HG orchard, and 0,150,300,450 g per tree (soil applied) or 50 g per tree (foliar) for the LG orchards. In all orchards the proportion of green colour on the ripe fruit was significantly (P<0.05) higher with soil applications of 150 g N or more per tree. Foliar sprays resulted in a higher proportion of green colour than the highest soil treatment in the HG orchard, but not in the LG orchards. Anthracnose disease severity was significantly (P<0.05) higher with 300 g of N per tree or foliar treatment in the HG orchard, compared with no additional N. Thus, N can reduce mango fruit quality by increasing green colour and anthracnose disease in ripe fruit.

Exposure of Tomato Fruit to 1-MCP Improves Quality of Stored Slices

Maintenance of quality, such as firmness, is important during storage of fresh cut produce. This study compared the effects of I-MCP on the quality of tomato slices when intact tomatoes were treated with 1-MCP and then sliced, or tomatoes were sliced and the slices treated with I-MCP. In both instances the MCP treatment was 1 µL Lˉٰ at 20 ºC for 12 h. Tomato cv. 'Revolution' was harvested at the 'pink' stage of maturity, cut into 7-mm slices, and stored as vertical stacks in closed plastic containers at 5ºC for up to 7 days after the 1-MCP treatment. Exposure of intact tomatoes to I-MCP resulted in reduced ethylene production (31%) and firmer (22%) slices than when tomatoes were not I-MCP treated. The application of I-MCP prior to slicing of tomatoes appears a useful strategy to retain quality of stored tomato slices.

Morphological Characteristics of Genetically-Modified Pineapple Fruit

Genetic engineering is an attractive method for changing a single characteristic of ‘Smooth Cayenne’ pineapple, without altering its other desirable attributes. Techniques used in pineapple transformation, however, such as tissue culture and biolistic-mediated or Agrobacterium-mediated gene insertion are prone to somaclonal variation, resulting in the production of several morphological mutations (Smith et al., 2002). Fruit mutations can include distortion in fruit shape (round ball, conical, fan-shaped), reduced fruit size, multiple crowns, crownless fruit, fruitless crowns, and spiny crown leaves (Dalldorf, 1975; Sanewski et al., 1992). The present paper describes the variability in fruit-shape mutations between transgenic and non-transgenic fruit, and its subsequent impact on organoleptic characteristics.

Pheromone-mediated mass trapping and population diversion as strategies for suppressing Carpophilus spp. (Coleoptera: Nitidulidae) in Australian stone fruit orchards

1 Five experiments were conducted during 1995-99 in stone fruit orchards on the Central Coast and in inland New South Wales, Australia, on the use of synthetic aggregation pheromones and a coattractant to suppress populations of the ripening fruit pests Carpophilus spp. (Coleoptera: Nitidulidae). 2 Perimeter-based suppression traps baited with pheromone and coattractant placed at 3m intervals around small fruit blocks, caught large numbers of Carpophilus spp. Very small populations of Carpophilus spp. occurred within blocks, and fruit damage was minimal. 3 Carpophilus spp. populations in stone fruit blocks 15-370m from suppression traps were also small and non-damaging, indicating a large zone of pheromone attractivity. 4 Pheromone/coattractant-baited suppression traps appeared to divert Carpophilus spp. from nearby (130 m) ripening stone fruit. Ten metal drums containing decomposing fruit, baited with pheromone and treated with insecticide, attracted Carpophilus spp. and appeared to reduce populations and damage to ripening fruit at distances of 200-500 m. Populations and damage were significantly greater within 200m of the drums and may have been caused by ineffective poisoning or poor quality/overcrowding of fruit resources in the drums. 5 Suppression of Carpophilus spp. populations using synthetic aggregation pheromones and a coattractant appears to be a realistic management option in stone fruit orchards. Pheromone-mediated diversion of beetle populations from ripening fruit may be more practical than perimeter trapping, but more research is needed on the effective range of Carpophilus pheromones and the relative merits of trapping compared to attraction to insecticide-treated areas.

Pheromone-trapping of Carpophilus spp. (Coleoptera: Nitidulidae) in stone fruit orchards near Gosford, New South Wales: Fauna, seasonality and effect of insecticides

Traps baited with synthetic aggregation pheromones of Carpophilus hemipterus (L.), Carpophilus mutilatus Erichson and Carpophilus davidsoni Dobson and fermenting bread dough were used to identify the fauna and monitor the seasonal abundance of Carpophilus spp. in insecticide treated peach and nectarine orchards in the Gosford area of coastal New South Wales. In four orchards 67 178 beetles were trapped during 1994–1995, with C. davidsoni (82%) and Carpophilus gaveni (Dobson) (12.2%) dominating catches. Five species (C. hemipterus, C. mutilatus, Carpophilus marginellus Motschulsky, Carpophilus humeralis (F.) and an unidentified species) each accounted for 0.2–3.2% of trapped beetles. Carpophilus davidsoni was most abundant during late September–early October but numbers declined rapidly during October, usually before insecticides were applied. Spring populations of Carpophilus spp. were very large in 1994–1995 (1843–2588 per trap per week). However, despite a preharvest population decline of approximately 95% and 2–11 applications of insecticide, 14–545 beetles per trap per week (above the arbitrary fruit damage threshold of 10 beetles per trap per week) were recorded during the harvest period and fruit damage occurred at three of the four orchards. Lower preharvest populations in 1995–1996 (< 600 per trap per week) and up to six applications of insecticide resulted in < 10 beetles per trap per week during most of the harvest period and minimal or no fruit damage. The implications of these results for the integrated management of Carpophilus spp. in coastal and inland areas of southeastern Australia are discussed.

Phenology of Carpophilus spp. (Coleoptera: Nitidulidae)in stone fruit orchards as determined by pheromone trapping: Implications for prediction of crop damage

Traps baited with synthetic aggregation pheromone and fermenting bread dough were used to monitor seasonal incidence and abundance of the ripening fruit pests, Carpophilus hemipterus (L.), C. mutilatus Erichson and C. davidsoni Dobson in stone fruit orchards in the Leeton district of southern New South Wales during five seasons (1991-96). Adult beetles were trapped from September-May, but abundance varied considerably between years with the amount of rainfall in December-January having a major influence on population size and damage potential during the canning peach harvest (late February-March). Below average rainfall in December-January was associated with mean trap catches of < 10 beetles/trap/week in low dose pheromone traps during the harvest period in 1991/92 and 1993/94 and no reported damage to ripening fruit. Rainfall in December-January 1992/93 was more than double the average and mean trap catches ranged from 8-27 beetles/week during the harvest period with substantial damage to the peach crop. December-January rainfall was also above average in 1994/95 and 1995/96 and means of 50-300 beetles/trap/week were recorded in high dose pheromone traps during harvest periods. Carpophilus spp. caused economic damage to peach crops in both seasons. These data indicate that it may be possible to predict the likelihood of Carpophilus beetle damage to ripening stone fruit in inland areas of southern Australia, by routine pheromone-based monitoring of beetle populations and summer temperatures and rainfall.

Assessment of internal quality attributes of mandarin fruit. 1. NIR calibration model development

The utility of near infrared spectroscopy as a non-invasive technique for the assessment of internal eating quality parameters of mandarin fruit (Citrus reticulata cv. Imperial) was assessed. The calibration procedure for the attributes of TSS (total soluble solids) and DM (dry matter) was optimised with respect to a reference sampling technique, scan averaging, spectral window, data pre-treatment (in terms of derivative treatment and scatter correction routine) and regression procedure. The recommended procedure involved sampling of an equatorial position on the fruit with 1 scan per spectrum, and modified partial least squares model development on a 720–950-nm window, pre-treated as first derivative absorbance data (gap size of 4 data points) with standard normal variance and detrend scatter correction. Calibration model performance for the attributes of TSS and DM content was encouraging (typical Rc2 of >0.75 and 0.90, respectively; typical root mean squared standard error of calibration of <0.4 and 0.6%, respectively), whereas that for juiciness and total acidity was unacceptable. The robustness of the TSS and DM calibrations across new populations of fruit is documented in a companion study.

Assessment of internal quality attributes of mandarin fruit. 2. NIR calibration model robustness

The robustness of multivariate calibration models, based on near infrared spectroscopy, for the assessment of total soluble solids (TSS) and dry matter (DM) of intact mandarin fruit (Citrus reticulata cv. Imperial) was assessed. TSS calibration model performance was validated in terms of prediction of populations of fruit not in the original population (different harvest days from a single tree, different harvest localities, different harvest seasons). Of these, calibration performance was most affected by validation across seasons (signal to noise statistic on root mean squared error of prediction of 3.8, compared with 20 and 13 for locality and harvest day, respectively). Procedures for sample selection from the validation population for addition to the calibration population (‘model updating’) were considered for both TSS and DM models. Random selection from the validation group worked as well as more sophisticated selection procedures, with approximately 20 samples required. Models that were developed using samples at a range of temperatures were robust in validation for TSS and DM.

Robustness of NIR calibrations for assessing fruit quality

Predictive models based on near infra-red spectroscopy for the assessment of fruit internal quality attributes must exhibit a degree of robustness across the parameters of variety, district and time to be of practical use in fruit grading. At the time this thesis was initiated, while there were a number of published reports on the development of near infra-red based calibration models for the assessment of internal quality attributes of intact fruit, there were no reports of the reliability ("robustness") of such models across time, cultivars or growing regions. As existing published reports varied in instrumentation employed, a re-analysis of existing data was not possible. An instrument platform, based on partial transmittance optics, a halogen light source and (Zeiss MMS 1) detector operating in the short wavelength near infra-red region was developed for use in the assessment of intact fruit. This platform was used to assess populations of macadamia kernels, melons and mandarin fruit for total soluble solids, dry matter and oil concentration. Calibration procedures were optimised and robustness assessed across growing areas, time of harvest, season and variety. In general, global modified partial least squares regression (MPLS) calibration models based on derivatised absorbance data were better than either multiple linear regression or `local' MPLS models in the prediction of independent validation populations . Robustness was most affected by growing season, relative to the growing district or variety . Various calibration updating procedures were evaluated in terms of calibration robustness. Random selection of samples from the validation population for addition to the calibration population was equivalent to or better than other methods of sample addition (methods based on the Mahalanobis distance of samples from either the centroid of the population or neighbourhood samples). In these exercises the global Mahalanobis distance (GH) was calculated using the scores and loadings from the calibration population on the independent validation population. In practice, it is recommended that model predictive performance be monitored in terms of predicted sample GH, with model updating using as few as 10 samples from the new population undertaken when the average GH value exceeds 1 .0 .

Discovery of genes associated with fruit ripening in arica papaya using expressed sequence tags

To identify genes involved in papaya fruit ripening, a total of 1171 expressed sequence tags (ESTs) were generated from randomly selected clones of two independent fruit cDNA libraries derived from yellow and red-fleshed fruit varieties. The most abundant sequences encoded: chitinase, 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase, catalase and methionine synthase, respectively. DNA sequence comparisons identified ESTs with significant similarity to genes associated with fruit softening, aroma and colour biosynthesis. Putative cell wall hydrolases, cell membrane hydrolases, and ethylene synthesis and regulation sequences were identified with predicted roles in fruit softening. Expressed papaya genes associated with fruit aroma included isoprenoid biosynthesis and shikimic acid pathway genes and proteins associated with acyl lipid catabolism. Putative fruit colour genes were identified due to their similarity with carotenoid and chlorophyll biosynthesis genes from other plant species.

Discovery of genes associated with fruit ripening in Carica papaya using expressed sequence tags

To identify genes involved in papaya fruit ripening, a total of 1171 expressed sequence tags (ESTs) were generated from randomly selected clones of two independent fruit cDNA libraries derived from yellow and red-fleshed fruit varieties. The most abundant sequences encoded:chitinase, 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase, catalase and methionine synthase, respectively. DNA sequence comparisons identified ESTs with significant similarity to genes associated with fruit softening, aroma and colour biosynthesis. Putative cell wall hydrolases, cell membrane hydrolases, and ethylene synthesis and regulation sequences were identified with predicted roles in fruit softening. Expressed papaya genes associated with fruit aroma included isoprenoid biosynthesis and shikimic acid pathway genes and proteins associated with acyl lipid catabolism. Putative fruit colour genes were identified due to their similarity with carotenoid and chlorophyll biosynthesis genes from other plant species.