Neural network based off-line handwritten text recognition system

This dissertation introduces a new system for handwritten text recognition based on an improved neural network design. Most of the existing neural networks treat mean square error function as the standard error function. The system as proposed in this dissertation utilizes the mean quartic error function, where the third and fourth derivatives are non-zero. Consequently, many improvements on the training methods were achieved. The training results are carefully assessed before and after the update. To evaluate the performance of a training system, there are three essential factors to be considered, and they are from high to low importance priority: (1) error rate on testing set, (2) processing time needed to recognize a segmented character and (3) the total training time and subsequently the total testing time. It is observed that bounded training methods accelerate the training process, while semi-third order training methods, next-minimal training methods, and preprocessing operations reduce the error rate on the testing set. Empirical observations suggest that two combinations of training methods are needed for different case character recognition. Since character segmentation is required for word and sentence recognition, this dissertation provides also an effective rule-based segmentation method, which is different from the conventional adaptive segmentation methods. Dictionary-based correction is utilized to correct mistakes resulting from the recognition and segmentation phases. The integration of the segmentation methods with the handwritten character recognition algorithm yielded an accuracy of 92% for lower case characters and 97% for upper case characters. In the testing phase, the database consists of 20,000 handwritten characters, with 10,000 for each case. The testing phase on the recognition 10,000 handwritten characters required 8.5 seconds in processing time.

Diel and seasonal variation in the use of a nearshore sandflat by a ray community in a near pristine system

Knowledge of movements and habitat use is necessary to assess a species’ ecological role and is especially important for mesopredators because they provide the link between upper and lower trophic levels. Using acoustic telemetry, we examined coarse-scale diel and seasonal movements of elasmobranch mesopredators on a shallow sandflat in Shark Bay, Western Australia. Giant shovelnose rays (Glaucostegus typus) and reticulate whiprays (Himantura uarnak) were most often detected in nearshore microhabitats and were regularly detected throughout the day and year, although reticulate whiprays tended to frequent the monitored array over longer periods. Pink whiprays (H. fai) and cowtail stingrays (Pastinachus atrus) were also detected throughout the day, but were far less frequently detected. Overall, there was no apparent spatial or temporal partitioning of the sandflats, but residency to the area varied between species. In addition, ray presence throughout the year suggests that previously observed differences in seasonal abundance are likely because of seasonal changes in habitat use rather than large-scale migrations. Continuous use of the sandflats and limited movements within this ray community suggests that rays have the potential to be a structuring force on this system and that focusing on nearshore habitats is important for managing subtropical ray populations.