Massey Documents by Type

Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294

Browse

Subject: search.filters.subject.Identification

Search Results

Now showing 1 - 10 of 13
  • Thumbnail Image
    Item
    Deep learning-based approaches for plant disease and weed detection : a thesis by publications presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Engineering, Massey University, Auckland, New Zealand
    (Massey University, 2022) Saleem, Muhammad Hammad
    To match the ever-growing food demand, the scientific community has been actively focusing on addressing the various challenges faced by the agricultural sector. The major challenges are soil infertility, abrupt changes in climatic conditions, scarcity of water, untrained labor, emission of greenhouse gases, and many others. Moreover, plant diseases and weeds are two of the most important agricultural problems that reduce crop yield. Therefore, accurate detection of plant diseases and weeds is one of the essential operations to apply targeted and timely control measures. As a result, this can improve crop productivity, reduce the environmental effects and financial losses resulting from the excessive application of fungicide/herbicide spray on diseased plants/weeds. Among various ways of plant disease and weed detection, image-based methods are significantly effective for the interpretation of the distinct features. In recent years, image-based deep learning (DL) techniques have been reported in literature for the recognition of weeds and plant diseases. However, the full potential of DL has not yet been explored as most of the methods rely on modifications of the DL models for well-known and readily available datasets. The current studies lack in several ways, such as addressing various complex agricultural conditions, exploring several aspects of DL, and providing a systematic DL-based approach. To address these research gaps, this thesis presents various DL-based methodologies and aims to improve the mean average precision (mAP) for the identification of diseases and weeds in several plant species. The research on plant disease recognition starts with a publicly available dataset called PlantVillage and comparative analyses are conducted on various DL feature extractors, meta-architectures, and optimization algorithms. Later, new datasets are generated from various local New Zealand horticultural farms, named NZDLPlantDisease-v1 & v2. The proposed datasets consist of healthy and diseased plant organs of 13 economically important horticultural crops of New Zealand, divided into 48 classes. A performance-optimized DL model and a transfer learning-based approach are proposed for the detection of plant diseases using curated datasets. The weed identification has been performed on an open-source dataset called DeepWeeds. A two-step weed detection pipeline is presented to show the performance improvement of the deep learning model with a significant margin. The results for both agricultural tasks achieve superior performance compared to the existing method/default settings. The research outcomes elaborate the practical aspects and extended potential of DL for selected agricultural applications. Therefore, this thesis is a benchmark step for cost-effective crop protection and site-specific weed management systems (SSWM).
  • Thumbnail Image
    Item
    Deep learning for asteroid detection in large astronomical surveys : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Computer Science at Massey University, Albany, New Zealand
    (Massey University, 2022) Cowan, Preeti
    The MOA-II telescope has been operating at the Mt John Observatory since 2004 as part of a Japan/NZ collaboration looking for microlensing events. The telescope has a total field of view of 1.6 x 1.3 degrees and surveys the Galactic Bulge several times each night. This makes it particularly good for observing short duration events. While it has been successful in discovering exoplanets, the full scientific potential of the data has not yet been realised. In particular, numerous known asteroids are hidden amongst the MOA data. These can be clearly seen upon visual inspection of selected images. There are also potentially many undiscovered asteroids captured by the telescope. As yet, no tool exists to effectively mine archival data from large astronomical surveys, such as MOA, for asteroids. The appeal of deep learning is in its ability to learn useful representations from data without significant hand-engineering, making it an excellent tool for asteroid detection. Supervised learning requires labelled datasets, which are also unavailable. The goal of this research is to develop datasets suitable for supervised learning and to apply several CNN-based techniques to identify asteroids in the MOA-II data. Asteroid tracklets can be clearly seen by combining all the observations on a given night and these tracklets form the basis of the dataset. Known asteroids were identified within the composite images, forming the seed dataset for supervised learning. These images were used to train several CNNs to classify images as either containing asteroids or not. The top five networks were then configured as an ensemble that achieved a recall of 97.67%. Next, the YOLO object detector was trained to localise asteroid tracklets, achieving a mean average precision (mAP) of 90.97%. These trained networks will be applied to 16 years of MOA archival data to find both known and unknown asteroids that have been observed by the telescope over the years. The methodologies developed can also be used by other surveys for asteroid recovery and discovery.
  • Thumbnail Image
    Item
    Genome-wide copy number variation in sheep : detection and utility as a genetic marker for quantitative traits, with reference to gastrointestinal nematodiasis : thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Animal Science at Massey University, Palmerston North, New Zealand
    (Massey University, 2018) Yan, Juncong
    Gastrointestinal nematodes are perhaps the most important parasites of domestic sheep world-wide. Genetic selection for nematode resistance in domestic sheep is being promoted in many countries including New Zealand. There are several strategies to identify genetic markers associated with quantitative traits. Single nucleotide polymorphism (SNP)-based strategies have been widely used in animal breeding. However, SNP cannot explain all the genetic variation for a particular trait. A new kind of variation, copy number variation (CNV) has been identified as contributing to genetic variation in production and disease traits. Compared with other domestic animals, CNV in sheep is poorly investigated. The primary objective of this thesis was to explore the utility of genome-wide CNV as a genetic marker for the analysis of quantitative traits in sheep. Five different studies were undertaken to fulfill the objective. The first two studies used 50 K SNP BeadChip genotype data and next generation sequencing (NGS) data to detect CNV. Extensive CNV differences were evident between breeds as well as detection algorithms. NGS-based detection resulted in better CNV resolution than that by SNP. Subsequently, a genome-wide association study (with a small sample size) using CNV detected from a high density (HD) SNP genotype data identified four CNV regions to be significantly associated with a couple of traits pertaining to gastrointestinal nematodiasis in Romney sheep, while no significant SNP associations were found. Somatic mosaicism of CNV, influenced by age (high in foetuses, compared to adults), individuals, detection algorithm and type of tissue analysed, was also evident in separate study. The final study detected CNV differences and SNP based selection signatures in two Romney lines selected for gastrointestinal nematode resistance or resilience. Several significant SNPs and line-specific CNV regions were identified. However, only one SNP overlapped to a CNV region, indicating that SNP-based selection signatures and CNV could represent different aspects of sheep immunogenetics. Overall, CNV could be a potential genetic marker, albeit with methods for detection and validation needing to be refined. The conclusions from this thesis expand our understanding of CNV in sheep and its potential application prospects for genetic breeding of sheep in the future.
  • Thumbnail Image
    Item
    Pangenome-guided tools for investigating the role of epsilonproteobacteria in human gastroenteritis : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy (PhD) in Veterinary Science at Massey University, Manawatu, New Zealand
    (Massey University, 2018) Cornelius, Angela Joyce
    Gastroenteritis affects billions of people every year and current testing methods fail to identify the cause for approximately half of the samples submitted for microbiological testing. Epsilonproteobacteria contains Campylobacter jejuni, the most commonly reported cause of bacterial gastroenteritis in the world, and Helicobacter pylori, a gastric pathogen and class I carcinogen. This bacterial class also contains ≥20 additional species known, or suspected, of being human pathogens. To better understand the role some of these species play in human gastroenteritis, novel rapid, cost effective methods are needed. The growing number of whole genome sequences available for this class were exploited to first evaluate the classification of the genetically heterogeneous species C. concisus and then to identify taxon-specific CDS for a range of Epsilonproteobacterial taxa. Probes were designed to detect 28 of these CDS and incorporated into a single multiplex ligation-dependent probe amplification (MLPA) assay which was tested against DNA from 43 Epsilonproteobacterial species and then applied to DNA extracts from stool samples from a childhood gastroenteritis case control study undertaken in Belgium. The 22 C. concisus genomes consistently clustered into two genomospecies (GS) represented by ATCC 33237T (GS1) and CCUG 19995 (GS2). Taxon-specific genes were identified for 28 taxa, including the two C. concisus genomospecies, and concordant results were observed for the majority of MLPA probes and DNA extracts from pure cultures. The probes designed to detect C. lari subsp. concheus and H. pullorum failed to detect the target DNA; all of the urease positive thermophilic Campylobacter DNA extracts were also positive for the probe designed to detect C. subantarcticus, some probes lacked repeatability in the presence of elevated EDTA and the size differences between some probes needs to be optimised. C. jejuni was the most common Epsilonproteobacterial species isolated by culture and C. concisus was the most common species detected by MLPA. Both C. jejuni and C. concisus GS2 were detected in significantly higher numbers in cases than controls in a Belgian childhood case control study. This demonstrated the utility of the Epsilonproteobacteria MLPA assay and provides some evidence that C. concisus GS2 may have a role in childhood gastroenteritis.
  • Thumbnail Image
    Item
    Development of a DNA hybridisation method for the identification of Rhizobium and Bradyrhizobium: a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in microbiology at Massey University, New Zealand
    (Massey University, 1990) Garman, Jean Heather
    The potential of a DNA hybridisation method, utilising a biotin-labelling system with a streptavidin/alkaline phosphatase detection system (ENZO Biochem), was investigated as an identification method for Rhizobium species and Bradyrhizobium sp. (Lotus) strains using nodule, colony and pure DNA. The method used for extracting DNA from colonies and crushed nodules and binding it to nitrocellulose did not purify the DNA sufficiently to stop non-specific binding occurring between the streptavidin-alkaline phosphatase conjugate and the sample. An alternative method of colony hybridisation that removed more of the cellular constituents was required. Only pure DNA could be used. The method was altered as follows: i) Tris/EDTA buffer was used to terminate the colour reaction in place of allowing the membrane to air dry; ii) 5% milk powder was used in place of 10% bovine serum albumin in the blocking buffer, complex detection buffer and washing buffer used in the detection of hybridised biotin-labelled DNA; iii) 5% dextran sulphate was included in the hybridisation buffer to decrease the minimum hybridisation time from 6hr to 3hr. Investigation of the effect of variable conditions on the intensity of colour produced showed that: i) the incubation of alkaline phosphatase with its substrate at room temperature resulted in fluctuation of the development time as the enzyme reaction rate is sensitive over this range of temperature (approximately 1s0 c to 30°C); ii) increasing the concentration of labelled DNA in the hybridisation buffer increased the intensity of colour produced, the minimum concentration that could be used without lowering the detection limit was 200 ng/ml; iii) continued incubation of alkaline phosphatase with its substrate after colour development in the negative control had begun gave an increased colour intensity in the sample but since this increase was not proportional to that of the negative control the net response (sample minus control) decreased. When genomic probes were hybridised with slot-blots containing homologous DNA the detection limit was between 63 and 125 ng of DNA. Both 32P-labelled and biotin­ labelled genomic Rhizobium leguminosarum biovar trifolii DNA probes were able to distinguish between Rhizobium leguminosarum and other Rhizobium species but not between the biovars of R.leguminosarum. To distinguish between closely related species or strains when using 32P-labelled or biotin-labelled probes a specific DNA sequence was required for use as the probe. Two distinct DNA homology groups have been described in Bradyrhizobium sp. (Lotus). From a gene library of Bradyrhizobium sp. (Lotus) strain cc814S (homology group I) 8 clones were isolated that contained sequences that distinguish a representative of homology group I (strain cc814S) from a representative of homology group II (strain NZP2076). This was achieved by hybridising total genomic DNA from strain cc814S with total genomic DNA from strain NZP2076 and removing the single stranded specific sequences with hydroxylapatite. The specific DNA was used to probe the gene library. Increased selection for group-specific sequences by substituting another homology group I strain (NZP2021) for strain cc814S and subcloning one of the clones isolated gave inconclusive results but indicated that a group specific sequence could be derived in this way.
  • Thumbnail Image
    Item
    Students with disability : data collection for reporting in New Zealand universities : a thesis presented in partial fulfillment of the requirements for the degree of Master of Philosophy in Rehabilitation at Massey University, Palmerston North, New Zealand
    (Massey University, 2003) Harris, Rosemary J
    The Ministry of Education in New Zealand has dedicated funding to increase the participation of people with disability in tertiary education. However there has been no standardised system in place for defining disability, categorising impairment, or collecting, maintaining and reporting data about tertiary students with disability, in order to determine the eventual impact of this initiative. The present study utilised a cross-sectional survey in a single stage sampling procedure, to gather information from the eight New Zealand universities regarding definitions of disability and categories of impairment used to collect data, as well as the source of data collection and numerical characteristics of the population. Data collected showed a steady increase in the population of university students with disability from 3,039 in 1998 to 4,358 in 2000. However the findings were consistent with the evidence in the literature review that it is currently not possible to know the real number of these students because of the differences in data collection and reporting across institutions. These findings indicated that data was not sourced in the same way across institutions. Furthermore, information was kept in segregated databases in some institutions, which did not all have a means of exchanging data with their general student record system. The Ministry of Education's reporting template introduced in 2001 was found to provide only a partially standardised framework for reporting on data. There must also be a systematic method of collecting and maintaining data across tertiary institutions, including clarification of the sets of students to be counted, so that all institutions are counting students in the same way. The present study identified confusion in language and definitions, with the terms impairment, disability, illness and injury being used interchangeably. The International Classification of Functioning, Disability and Health was suggested as providing a practical functional model for data collection, which could be used as a platform for establishing definitions and clarifying the language around disability and impairment, as well as providing an international standard for establishing consistency.
  • Thumbnail Image
    Item
    Development of a low-cost automated sample presentation and analysis system for counting and classifying nematode eggs : a thesis presented in partial fulfilment of the requirements for the degree of Master of Engineering in Mechatronics at Massey University, Manawatu, New Zealand
    (Massey University, 2017) Pedersen, Benjamin
    This thesis discusses the concept development and design of a low-cost, automated, sample presentation system for faecal egg counting, and classification. The system developed uses microfluidics to present nematode eggs for digital imaging to produce images suitable for image analysis and classification. The system costs are kept low by using simple manufacturing methods and commonly available equipment to produce microfluidic counting chambers, which can be interfaced with conventional microscopes. This thesis includes details of the design and implementation of the software developed to allow capture and processing of images from the presentation system. This thesis also includes details on the measures taken to correct for the optical aberrations introduced by the sample presentation system.
  • Thumbnail Image
    Item
    The casemoth, Liothula omnivoa (Psychidae : lepidoptera) : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Zoology at Massey University
    (Massey University, 1967) Ooi, Thean Chooi
    Liothula omnivora, one of the two known casemoths endemic to New Zealand, belongs to the Lepidopteran family Psychidae. It is distributed throughout the country, and can be found on a large number of host plants (see later). The other N.Z. casemoth, Orophora concolor, has been found on Wild Irishman and cassinias in the river beds of the South Island (Miller, 1955). L. omnivora was first described by Fereday in 1878, but Meyrick (1890) transferred it to the genus Oiketicus (Guilding, 1827) mis-spelling it Oeceticus. Dr. Allan Watson (1967, pers. comm.) of the British Museum (Natural History) considers that this species should belong in the genus Liothula and the writer has adopted Watson's view in calling it L. omnivora. The type of L. omnivora is in the Canterbury Museum, Christchurch (Entomologische Beihefte 4, Horn and Kahle, 1937). Descriptions of the external morphology of the adult male and female have been made by Fereday (1878), Meyrick (1890) and Hudson (1928). Fereday and Hudson also described the larva, the pupa has been described by Hudson and Quail (1901), and the appearance of the egg briefly noted by Hudson.
  • Thumbnail Image
    Item
    The development of diagnostic tools for the grapevine pathogen Eutypa lata : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Genetics at Massey University
    (Massey University, 2001) Jones, Paula Elizabeth
    Eutypa lata is the causal agent of Eutypa dieback on grapevines. The fungus invades the vine and grows there unnoticed, possibly for several years, causing discolouration and deformation of the vine shoots and leaves. Most berries fail to establish on these shoots and the fungus eventually kills the vine. The damaging effects of this fungus have had a notable financial impact on the grape and wine industry world wide and E. lata is at present the primary constraint on vineyard longevity in many places including California and Australia. Little is known about the occurrence and distribution of Eutypa dieback within New Zealand. This is due mainly to difficulties associated with identification of the disease in grapevines. To develop a molecular probe for the identification of E. lata from grapevine wood the Polymerase Chain Reaction (PCR) amplified the Internal Transcribed Spacers (ITS1 and ITS2) and the intervening 5.8S gene of ribosomal DNA (rDNA) from representative isolates. The sequences of the E. lata ITS regions were used to design two pairs of primers, each of which was subsequently shown to be specific for the amplification of predicted-size fragments from genomic DNA of E. lata. The primer pairs were further tested using template DNA extracted from healthy grapevines and from other fungi commonly isolated from dieback diseased grapevines but no PCR amplification was observed. Simple DNA extraction protocols, leading to the rapid release of DNA, were tested to enable identification of E. lata from pure culture and grapevine wood; however, a suitable DNA extraction method from these materials was not found. Currently the only known source of inoculum is ascospores, which are released from perithecia during and immediately after rainfall. However, few perithecia have been found in New Zealand vineyards. This has prompted the study of the mating habits of E. lata. As the sexual stage of E. lata cannot be obtained in culture at present, the analysis of its mating system must be performed in natural populations. Molecular characterisation of the mating type at the outset of a mating project allows significant savings in time and effort as it drastically reduces the number of crosses that must be set up. So far, cloning of mating type (MAT) genes from fungi has been hampered by low conservation among them. Most ascomycete fungi have one mating type gene with two alternative forms or idiomorphs (MAT1-1 and MAT1-2). One of the pair of MAT genes. MAT1-2, encodes a protein with a conserved DNA binding motif called the high mobility group (HMG) box. There is sufficient sequence conservation at the borders of the HMG box to allow PCR amplification. New Zealand isolates of E. lata, including sixteen single ascospore isolates from one perithecium, were tested for the presence of a MAT1-2 idiomorph using this PCR based approach. Five different sets of primers were used which were designed to anneal at different target sites with different specificities. PCR products of the expected size were obtained and sequenced, but despite exhaustive attempts to optimise PCR specificity, none of these had convincing homology to fungal mating type genes. Progress on the basic aspects of the genetics of E. lata will continue to be hampered until the organism is induced to complete its life cycle in culture. Molecular studies into the mating type genes which regulate sexual compatibility and sexual reproduction in the fungus should lead to a deeper understanding of the life-cycle of E. lata and the critical influence of sex on population genetics. In addition, it will provide a scientific basis for a management program urgently needed to minimise the impact of this disease.
  • Thumbnail Image
    Item
    Identifying giftedness in early childhood centres : a thesis presented in partial fulfilment of the requirements for the degree of Master of Education at Massey University, Palmerston North, New Zealand
    (Massey University, 1999) Allan, Barbara Ann
    This study investigated current understanding of giftedness as it relates to NZ early childhood centre settings, in order to produce a teacher-friendly identification tool and to explore the effect of identification on curriculum provision for young children displaying gifted behaviours. Analysis of international research literature provided an initial source of indicators that could be used by teachers within the specific context of NZ early childhood centres in order to identify gifted behaviours in young children. Academics involved in gifted education and early childhood teachers experienced with gifted young children critiqued an identification instrument based on these indicators. Modifications based on these critiques resulted in an instrument of indicators of gifted behaviours considered relevant to NZ early childhood settings grouped under headings of cognition and language, approach to learning, creativity and social competence. Seventeen early childhood centres, involving a total of 167 children selected on the basis of age, gender, and ethnicity only, trialled the instrument. Seven centres participated in a training workshop previous to trialling the instrument, 10 centres received no pre-trial training. Focus group interviews revealed that using the instrument increased teachers' understanding and recognition of gifted behaviour, but that participation in a short training session did not increase success in identifying giftedness. Teachers did not show clear understanding of giftedness relating to diverse cultures or negative behaviour. A further phase of the research used unstructured interviews in six individual centres over one month to investigate the impact of identification on provision for gifted children. Teachers expressed a need for support services to assist in catering for gifted young children. The research demonstrated that while the identification instrument was useful to teachers, there are needs for further professional support and extended preservice and in-service training regarding both the diversity of giftedness and the provision of differentiated programmes for gifted young children.