Massey Documents by Type

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

Browse

Filters

2020 - 20222

Settings

Has files: YesSubject: search.filters.subject.460306 Image processing

Search Results

Now showing 1 - 2 of 2
  • 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
    Sketch recognition of digital ink diagrams : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Computer Science at Massey University, Palmerston North, New Zealand
    (Massey University, 2020) Ghodrati, Amirhossein
    Sketch recognition of digital ink diagrams is the process of automatically identifying hand-drawn elements in a diagram. This research focuses on the simultaneous grouping and recognition of shapes in digital ink diagrams. In order to recognise a shape, we need to group strokes belonging to a shape, however, strokes cannot be grouped until the shape is identified. Therefore, we treat grouping and recognition as a simultaneous task. Our grouping technique uses spatial proximity to hypothesise shape candidates. Many of the hypothesised shape candidates are invalid, therefore we need a way to reject them. We present a novel rejection technique based on novelty detection. The rejection method uses proximity measures to validate a shape candidate. In addition, we investigate on improving the accuracy of the current shape recogniser by adding extra features. We also present a novel connector recognition system that localises connector heads around recognised shapes. We perform a full comparative study on two datasets. The results show that our approach is significantly more accurate in finding shapes and faster on process diagram compared to Stahovich et al. (2014), which the results show the superiority of our approach in terms of computation time and accuracy. Furthermore, we evaluate our system on two public datasets and compare our results with other approaches reported in the literature that have used these dataset. The results show that our approach is more accurate in finding and recognising the shapes in the FC dataset (by finding and recognising 91.7% of the shapes) compared to the reported results in the literature.