Semantic segmentation of astronomical radio images: a computer vision approach
- Authors: Kupa, Ramadimetse Sydil
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422378 , vital:71937
- Description: The new generation of radio telescopes, such as the MeerKAT, ASKAP (Australian Square Kilometre Array Pathfinder) and the future Square Kilometre Array (SKA), are expected to produce vast amounts of data and images in the petabyte region. Therefore, the amount of incoming data at a specific point in time will overwhelm any current traditional data analysis method being deployed. Deep learning architectures have been applied in many fields, such as, in computer vision, machine vision, natural language processing, social network filtering, speech recognition, machine translation, bioinformatics, medical image analysis, and board game programs. They have produced results which are comparable to human expert performance. Hence, it is appealing to apply it to radio astronomy data. Image segmentation is one such area where deep learning techniques are prominent. The images from the new generation of telescopes have a high density of radio sources, making it difficult to classify the sources in the image. Identifying and segmenting sources from radio images is a pre-processing step that occurs before sources are put into different classes. There is thus a need for automatic segmentation of the sources from the images before they can be classified. This work uses the Unet architecture (originally developed for biomedical image segmentation) to segment radio sources from radio astronomical images with 99.8% accuracy. After segmenting the sources we use OpenCV tools to detect the sources on the mask images, then the detection is translated to the original image where borders are drawn around each detected source. This process automates and simplifies the pre-processing of images for classification tools and any other post-processing tool that requires a specific source as an input. , Thesis (MSc) -- Faculty of Science, Physics and Electronics, 2023
- Full Text:
- Authors: Kupa, Ramadimetse Sydil
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422378 , vital:71937
- Description: The new generation of radio telescopes, such as the MeerKAT, ASKAP (Australian Square Kilometre Array Pathfinder) and the future Square Kilometre Array (SKA), are expected to produce vast amounts of data and images in the petabyte region. Therefore, the amount of incoming data at a specific point in time will overwhelm any current traditional data analysis method being deployed. Deep learning architectures have been applied in many fields, such as, in computer vision, machine vision, natural language processing, social network filtering, speech recognition, machine translation, bioinformatics, medical image analysis, and board game programs. They have produced results which are comparable to human expert performance. Hence, it is appealing to apply it to radio astronomy data. Image segmentation is one such area where deep learning techniques are prominent. The images from the new generation of telescopes have a high density of radio sources, making it difficult to classify the sources in the image. Identifying and segmenting sources from radio images is a pre-processing step that occurs before sources are put into different classes. There is thus a need for automatic segmentation of the sources from the images before they can be classified. This work uses the Unet architecture (originally developed for biomedical image segmentation) to segment radio sources from radio astronomical images with 99.8% accuracy. After segmenting the sources we use OpenCV tools to detect the sources on the mask images, then the detection is translated to the original image where borders are drawn around each detected source. This process automates and simplifies the pre-processing of images for classification tools and any other post-processing tool that requires a specific source as an input. , Thesis (MSc) -- Faculty of Science, Physics and Electronics, 2023
- Full Text:
Source of introduced populations of Hemidactylus mabouia (Moreau de Jonnès, 1818) into the Eastern Cape Province and the potential spread to other South African regions
- Authors: Jeme, Alungile
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422334 , vital:71933
- Description: Human movements globally have resulted in dispersal and introduction of terrestrial and marine organisms into areas outside of their native range. Species that have been introduced to new habitats can either be alien (not causing any harm to the new environment) or invasive alien species (destroying or competing with the indigenous species for resources). Investigating the source of these introductions is important as it provides baseline information about their biology and evolution. It also allows for better prevention measures for future invasions and for effective conservation strategies. In this study, I investigate the source of the introduced populations of tropical house geckos Hemidactylus mabouia in the Eastern Cape Province using two mitochondrial gene regions, ribosomal 16S and NADH dehydrogenase subunit 4 (ND4). We also look at the potential spread of these species using species distribution models (SDMs). The phylogeny showed that the introduced populations into the Eastern Cape were genetically similar to both H. mabouia lineages found in Central & Southern Africa. The phylogeny produced two clades from which the majority of the introduced samples were grouped in a clade with samples from Angola, DRC, and Mozambique, and only a few nested within the second clade with the South African native population. Haplotype networks from both genes also showed two clades, matching those found in the phylogeny. The introduced samples were associated with samples from Angola and those from South Africa (native) as these localities had a number of shared haplotypes. SDMs showed less evidence that these individuals could naturally move west down the coast as the probability of occurrence was below 0.4 from the Eastern Cape to the Western Cape. Our conclusion was that the introductions may be due to multiple avenues like species moving down the east coast or human translocation including trade routes. From this we can conclude that geckos are largely translocated through human movement as they are known for their opportunistic behaviours as they are commensal with humans. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2023
- Full Text:
- Authors: Jeme, Alungile
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422334 , vital:71933
- Description: Human movements globally have resulted in dispersal and introduction of terrestrial and marine organisms into areas outside of their native range. Species that have been introduced to new habitats can either be alien (not causing any harm to the new environment) or invasive alien species (destroying or competing with the indigenous species for resources). Investigating the source of these introductions is important as it provides baseline information about their biology and evolution. It also allows for better prevention measures for future invasions and for effective conservation strategies. In this study, I investigate the source of the introduced populations of tropical house geckos Hemidactylus mabouia in the Eastern Cape Province using two mitochondrial gene regions, ribosomal 16S and NADH dehydrogenase subunit 4 (ND4). We also look at the potential spread of these species using species distribution models (SDMs). The phylogeny showed that the introduced populations into the Eastern Cape were genetically similar to both H. mabouia lineages found in Central & Southern Africa. The phylogeny produced two clades from which the majority of the introduced samples were grouped in a clade with samples from Angola, DRC, and Mozambique, and only a few nested within the second clade with the South African native population. Haplotype networks from both genes also showed two clades, matching those found in the phylogeny. The introduced samples were associated with samples from Angola and those from South Africa (native) as these localities had a number of shared haplotypes. SDMs showed less evidence that these individuals could naturally move west down the coast as the probability of occurrence was below 0.4 from the Eastern Cape to the Western Cape. Our conclusion was that the introductions may be due to multiple avenues like species moving down the east coast or human translocation including trade routes. From this we can conclude that geckos are largely translocated through human movement as they are known for their opportunistic behaviours as they are commensal with humans. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2023
- Full Text:
The development of a low-cost real- time bacterial plate culture monitoring system
- Authors: Songcata, Ntobeko
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422437 , vital:71942
- Description: Access restricted. Access embargoed until 2025. , Thesis (MSc) -- Faculty of Science, Biotechnology Innovation Centre, 2023
- Full Text:
- Authors: Songcata, Ntobeko
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422437 , vital:71942
- Description: Access restricted. Access embargoed until 2025. , Thesis (MSc) -- Faculty of Science, Biotechnology Innovation Centre, 2023
- Full Text:
The influence of immuno-contraception on the ecology of African savanna elephants (Loxodonta africana) in a small game reserve in the Eastern Cape, South Africa
- Authors: Fowler, Natalie
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422269 , vital:71926
- Description: African savanna elephants (Loxodonta africana) can alter the vegetation structure on a landscape which has both positive and negative consequences for biodiversity. This ecological role can be dependent on the density of elephants in a given area and how the area is transformed. Elephants are currently listed as an endangered species on the International Union for Conservation of Nature (IUCN) global red list. Despite this endangered status, strong conservation efforts in southern Africa have increased elephant numbers in protected areas. In South Africa, many of these protected areas are small (<1000 km2) and fenced. Thus, without the freedom to roam over large areas, elephant impacts on vegetation and other ecosystem processes may be exacerbated. Increased growth rates and higher population densities, have prompted wildlife managers to explore several population-level management techniques to slow elephant growth rates in small, protected areas. One such technique is the use of immuno-contraception. My study focused on three datasets (Pre, During and Post porcine zona pellucida (PZP) treatments) of one elephant population over an 18-year period in a small game reserve in the Eastern Cape, South Africa. The first aim of my study was to compare the female demography and reproductive parameters of treated (with PZP) and non-treated females (no PZP). I predicted that the population growth would decrease as more elephant cows were contracepted as well as them having longer inter-calving intervals than non-treated cows. Further, I predicted that there would be a skewed sex ratio of male calves born to treated cows. Lastly, I predicted that fertility rates would be lower in treated cows than in non-treated cows. I calculated the annual population growth rates, inter-calving intervals, sex ratios, age at first calving, fecundity rates and fertility rates of all breeding cows. Population growth rates were similar to other managed populations although, compared to a relatively untouched elephant population, annual growth rates in my study site were high. The inter-calving intervals and age at first birth were low compared to other African protected areas which were most likely influenced by the use of PZP, rainfall and a lack of density dependence. Treated cows did not have longer inter-calving intervals compared to non-treated cows nor was there a male biased sex ratio of the offspring of treated cows. The second aim was to assess the spatial ecology of this elephant population and how it may be influenced by immuno-contraception. I predicted that the use of PZP would increase the elephant herd’s core range as there were fewer reproducing cows. I also predicted that PZP would alter the elephant’s habitat use making them less selective in terms of the vegetation types they selected. I compared the home ranges of the elephant population among the three treatments (Pre PZP, During PZP and Post PZP treatment). Home ranges were larger when elephants were treated with PZP compared to the Pre and Post treatments. The herd utilized most of the reserve during the use of PZP whereas prior to contraception, their space use was mostly concentrated on the eastern side of the reserve. Additionally, the vegetation utilization was inconsistent across the differing timeframes, specifically between pre PZP and during PZP. Prior to the use of PZP, the elephant herd utilized the old lands and riverine thicket the most whereas during the use of PZP, the grasslands and karroid habitats were used the most. The most likely influence of the changes in home ranges and vegetation utilization were rainfall, the use of PZP and resource availability. My results suggest that the use of PZP is essential for controlling elephant population numbers. However, there may be some long-term effects (e.g., behavioural changes) of the use of PZP in elephant populations in small game reserves. Thus, I recommend continued monitoring of all aspects of elephant ecology on reserves where PZP is used as a contraceptive technique. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2023
- Full Text:
- Authors: Fowler, Natalie
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422269 , vital:71926
- Description: African savanna elephants (Loxodonta africana) can alter the vegetation structure on a landscape which has both positive and negative consequences for biodiversity. This ecological role can be dependent on the density of elephants in a given area and how the area is transformed. Elephants are currently listed as an endangered species on the International Union for Conservation of Nature (IUCN) global red list. Despite this endangered status, strong conservation efforts in southern Africa have increased elephant numbers in protected areas. In South Africa, many of these protected areas are small (<1000 km2) and fenced. Thus, without the freedom to roam over large areas, elephant impacts on vegetation and other ecosystem processes may be exacerbated. Increased growth rates and higher population densities, have prompted wildlife managers to explore several population-level management techniques to slow elephant growth rates in small, protected areas. One such technique is the use of immuno-contraception. My study focused on three datasets (Pre, During and Post porcine zona pellucida (PZP) treatments) of one elephant population over an 18-year period in a small game reserve in the Eastern Cape, South Africa. The first aim of my study was to compare the female demography and reproductive parameters of treated (with PZP) and non-treated females (no PZP). I predicted that the population growth would decrease as more elephant cows were contracepted as well as them having longer inter-calving intervals than non-treated cows. Further, I predicted that there would be a skewed sex ratio of male calves born to treated cows. Lastly, I predicted that fertility rates would be lower in treated cows than in non-treated cows. I calculated the annual population growth rates, inter-calving intervals, sex ratios, age at first calving, fecundity rates and fertility rates of all breeding cows. Population growth rates were similar to other managed populations although, compared to a relatively untouched elephant population, annual growth rates in my study site were high. The inter-calving intervals and age at first birth were low compared to other African protected areas which were most likely influenced by the use of PZP, rainfall and a lack of density dependence. Treated cows did not have longer inter-calving intervals compared to non-treated cows nor was there a male biased sex ratio of the offspring of treated cows. The second aim was to assess the spatial ecology of this elephant population and how it may be influenced by immuno-contraception. I predicted that the use of PZP would increase the elephant herd’s core range as there were fewer reproducing cows. I also predicted that PZP would alter the elephant’s habitat use making them less selective in terms of the vegetation types they selected. I compared the home ranges of the elephant population among the three treatments (Pre PZP, During PZP and Post PZP treatment). Home ranges were larger when elephants were treated with PZP compared to the Pre and Post treatments. The herd utilized most of the reserve during the use of PZP whereas prior to contraception, their space use was mostly concentrated on the eastern side of the reserve. Additionally, the vegetation utilization was inconsistent across the differing timeframes, specifically between pre PZP and during PZP. Prior to the use of PZP, the elephant herd utilized the old lands and riverine thicket the most whereas during the use of PZP, the grasslands and karroid habitats were used the most. The most likely influence of the changes in home ranges and vegetation utilization were rainfall, the use of PZP and resource availability. My results suggest that the use of PZP is essential for controlling elephant population numbers. However, there may be some long-term effects (e.g., behavioural changes) of the use of PZP in elephant populations in small game reserves. Thus, I recommend continued monitoring of all aspects of elephant ecology on reserves where PZP is used as a contraceptive technique. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2023
- Full Text:
The proprietary consequences of foreign marriages for the purposes of estate planning and succession in South Africa
- Authors: Thorne, Aimee Liza
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/420725 , vital:71773
- Description: This thesis is intended to provide clarity regarding the proprietary consequences of foreign marriages in the context of estate planning and succession in South Africa. It includes an exploration into the impact of specific legislation on foreign marriages, the implications of matrimonial property regimes on a deceased estate in South Africa, private international law principles affecting matrimonial property and succession claims, the scope of antenuptial contracts and postnuptial contracts as existing remedies for foreign remedies, and the harmonisation of succession and matrimonial property law in the European Union (EU). It furthermore suggests potential reforms for the overhaul of the lex domicilii matrimonii principle and associated issues arising therefrom. Various specific issues discussed include the waiver of accrual claims, maintenance claims by surviving spouses of a foreign marriage, choice-of-law rules for matrimonial property rights and succession rights, the registration of a domestic or foreign antenuptial contract for foreign marriages, the formal validity and proper law of antenuptial contracts, the operation of the immutability principle on a domestic and international level, and the common-law approach to the recognition of postnuptial contracts and spousal donations. , Thesis (LLM) -- Faculty of Law, Law, 2023
- Full Text:
- Authors: Thorne, Aimee Liza
- Date: 2023-03-29
- Subjects: Uncatalogued
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/420725 , vital:71773
- Description: This thesis is intended to provide clarity regarding the proprietary consequences of foreign marriages in the context of estate planning and succession in South Africa. It includes an exploration into the impact of specific legislation on foreign marriages, the implications of matrimonial property regimes on a deceased estate in South Africa, private international law principles affecting matrimonial property and succession claims, the scope of antenuptial contracts and postnuptial contracts as existing remedies for foreign remedies, and the harmonisation of succession and matrimonial property law in the European Union (EU). It furthermore suggests potential reforms for the overhaul of the lex domicilii matrimonii principle and associated issues arising therefrom. Various specific issues discussed include the waiver of accrual claims, maintenance claims by surviving spouses of a foreign marriage, choice-of-law rules for matrimonial property rights and succession rights, the registration of a domestic or foreign antenuptial contract for foreign marriages, the formal validity and proper law of antenuptial contracts, the operation of the immutability principle on a domestic and international level, and the common-law approach to the recognition of postnuptial contracts and spousal donations. , Thesis (LLM) -- Faculty of Law, Law, 2023
- Full Text: