Doctor of Philosophy (Ph.D.) Theses and Dissertations
Permanent URI for this collection
Browse
Recent Submissions
Now showing 1 - 5 of 16
- ItemOpen AccessDevelopment of a Segmentation-Based Deformation-Invariant face Recognition model(The Department of Computer Science and Engineering, Faculty of Technology, Obafemi Awolowo University., 2022) ALABI, Akeem AdisaFacial deformation has been a prominent issue in today’s trend of face recognition being a key product of human’s most frequently observed phenomenon called expression. The fact is that existing models are yet to fully capture aspects of deformation beyond shape and size of the face thus calling for new approaches to improve on present recognition models. This study formulated, implemented and evaluated a segmentation-based model with the view to recognising faces with deformities. A database of face images with different forms of deformation was created by collecting ten (10) different photographs of twenty (20) persons using digital camera and was analysed by discussing their key components. The collected images were grouped into training set and test data. These images were well cropped and then split accordingly through segmentation technique making each of the image features an individual image. The system model was formulated using modified eigenface algorithm incorporated with a three-phase verification key. This gave rise to the introduction of the Aggregated threshold (AT) as against the Uniform Threshold (UT) as the main parameter for the validation of results of comparison during detection and identification of the test images. The system was implemented using Open Computer Vision (i.e. Open CV) with Python programming language. The system evaluation was carried out to determine the role of the Aggregated Threshold as regards to the performance of the model. The evaluation was also extended to determine the behaviour of the system vis-à-vis the change in the pixel evaluation of the set of images. The results obtained showed that the performance of the proposed model outweighed the existing models as far as recognising the test images was concerned. To be precise, all the forty (40) test images were recognised in this model as opposed to the result in the existing model xvi where only 32 out of 40 images were recognized. This represents an increment of 20% accuracy recorded in comparison with the existing model. The use of the Aggregated Threshold through segmentation paved way for harnessing more information from the set of images during training culminating into the successes recorded in this study. It was also observed that the results obtained were similar with little discrepancies in the execution over the range 100x100 down to 50x50 dimensions. In other words, the number of identified test images remained the same with repeated execution of the code. The study concluded that the segmentation procedure introduced in this model gave rise to an enhanced system in the recognition of faces with deformities thereby giving individuals with this problem an opportunity to be recognized by a robust model.
- ItemEmbargoDevelopment of a validated dataset and a framework to mitigate bias in facial image processing(Department of Computer Science And Engineering, Faculty of Technology, Obafemi Awolowo University, Ile-Ife., 2025) Amarachi, Modester Udefi.This study demonstrated the levels of bias in facial image processing arising from a dataset, built a facial image dataset representing the biased population, and formulated an expression and gender recognition model to validate the dataset. It also described a framework showing the needed representation of certain demographic groups to mitigate bias in facial image processing. The performance of the dataset and model were also evaluated. These were with a view to developing a validated dataset and a framework to mitigate bias in facial image processing. A comprehensive review of 40 publicly accessible facial image datasets was conducted. To visualize the racial distribution of the datasets, t-distributed Stochastic Neighbor Embedding (t-SNE) was employed. Oriented FAST and Rotated BRIEF (ORB) were utilized for feature extraction, followed by K-means clustering to group racial features and Principal Component Analysis (PCA) to assess the geo-diversity and bias levels of the datasets. A 64MPX Camera was used to capture facial images in a controlled environment while questionnaires were used to gather the ground truths. A standard labeling convention was employed in labeling the dataset such that each participant was assigned a unique identifier: a string of ten characters as 0001DMY30C. Expression and Gender recognition models were developed using a Convolutional Neural Network Architecture in conjunction with a transfer learning technique. The UTK (University of Tennessee, Knoxville) dataset was used to train machine learning models to establish a framework to mitigate dataset bias. The model was evaluated based on accuracy, precision, and sensitivity metrics, while fairness metrics, such as demographic parity and equalized odds, were used to assess and quantify biases in the framework. From the result obtained, the PCA and k-means algorithms successfully identified the degree of bias in facial image datasets used in the analysis. The PCA also gave a visual representation of the bias levels in the form of scattered plots and bi-plots, where the facial image datasets were distinguished by their bias levels. A total of 3500 facial expression images were collected and used to develop a gender and expression recognition model. The gender recognition model presented an accuracy, precision, and sensitivity of 94%, 94%, and 94%, respectively, while the expression recognition model showed an accuracy, precision, and sensitivity of 96%, 90%, and 90%. The accuracy evaluation performance matrix for each ethnicity: Black, White, Latino, Asian, Indian, and Others is 98%, 92%, 88%, 89%, 88%, and 84%, respectively. The study concluded that the developed validated dataset and the framework were adequate and could be used to mitigate dataset bias in facial image processing. The framework effectively utilized a class weight formula to combat bias.
- ItemOpen AccessModeling and optimization of biotransformation of benzaldehyde to L-Phenylaceylcarbinol (L-PAC) by free cells of Torulaspora delbrueckii in the presence of B-Cyclodextrin(Department of Chemical Engineering , Faculty of Technology, Obafemi Awolowo University., 2015) Adepoju, Tunde FolorunsoIn this work, modeling and optimization of biotransformation of benzaldehde to L-phenlacetylcarbinol (L-PAC) using free cells of Torulaspora delbueckii in the presence of B-cyclodextrin was carried out. The model design was optimized using the Response Surface Methodology (RSM) and Artificial Neural Network (ANN). Furthermore, investigation of the effect of design variables on yield of biotransformation products.
- ItemOpen AccessDevelopment of Tea-Bag Product from Selected Costus Species in South West Nigeria for Anti-Hyperlycemic Activity.(Department of Chemical Engineering, Faculty of Technology, Obafemi Awolowo University, Ile-Ife, Nigeria., 2023) JIBOKU, Idowu BabatundeThe study carried out morphological and molecular characterization of selected Costus species in Southwestern Nigeria. It optimized extraction parameters using various solvents for Costus species extracts, evaluated the anti-diabetic activity of the selected Costus Species extract, and determined the toxicological effect on the most active extract. The chemical constituents of the most active extract were identified. The study also determined the anti-hyperglycemic effect of the selected Costus extracts formulated into tea bags. These were with a view to developing an affordable management for diabetes with a regulated therapeutic dosage from Costus species. The leaves of the Costus species were collected and identified using morphological features while molecular characterization was carried out using DNA extraction, Gel Electrophoresis, PCR, sequencing, and bioinformatics. The leaves were dried, pulverized, and extracted with n-Hexane, methanol, and water and the resulting filtrates were concentrated to obtain the crude extracts. Acute toxicity tests of the plant extracts were assessed. The effects of the extract on haematological and biochemical parameters were evaluated using standard procedures. Its antidiabetic activities were assayed in glucose and streptozotocin-induced diabetic rats’ models at various doses with glibenclamide. Histopathological examination of the pancreas, liver and kidney of rats administered with the extract at 250, 500 and 1000 mg/kg for 21 days were also carried out. The phytochemical components of the extract were analysed with LC-MS. The results obtained from these studies were subjected to statistical analysis using Analysis of Variance (ANOVA), followed by the Student Newman Keul's test. The morphological initially identified the species as C. igneus, C. afer, and C .dubius while molecular characterization confirmed them as C. pictus, Costus pulverentus, and Costus dubius respectively. Optimal conditions for C. igneus xxii extraction were 25.655 g/mL leaf-to-solvent ratio, 40.26 h extraction time, using methanol, yielding 12.76 wt.%. C. afer extraction's optimal conditions were 30 g/mL leaf-to-solvent ratio, 12 h extraction time, using methanol, with a yield of 7.20 wt.%. For C. dubius, optimal conditions were 25.6551 g/mL leaf-to-solvent ratio, 40.26 h extraction time, and methanol, yielding 13 wt.%. The median lethal dose, LD50 of the extracts was above 5000 mg/kg in rats. The extract of C. dubius elicited the highest antihyperglycaemic effect among the three Costus spp in glucose–induced hyperglycaemic study while its methanol and aqueous extracts effectively reduced hyperglycaemia in streptozotocin-induced diabetes. The aqueous extract of C. dubius at 250-1000 mg/kg significantly reduced PCV, WBC and haemoglobin levels of the rats and potentiated platelet levels. It also caused a significant increase in cholesterol, AST and ALT; levels. In addition, it caused various levels of distortion to the cytoarchitecture of the pancreas, liver and kidney at 250-1000 mg/kg. LC-MS analysis of the extract of C. dubius showed the presence of tannins, flavonoids, saponins, alkaloids and terpenoids The study concluded that the extract C. dubius had a potent antidiabetic effect with adverse heamatological, biochemical effects as well as toxic effect on the pancreas, liver and kidney. This called for caution in its use and formulation into a tea bag as an antidiabetic agent.
- ItemOpen AccessDevelopment of an Adsorption Solar Drying with Internet of Things- Based Control System.(Department of Agricultural and Environmental Engineering, Faculty of Technology, Obafemi Awolowo University., 2023) Olagunju, Titilope ModupeThis study investigated the effect of different adsorbent filters on the relative humidity of air and selected the most appropriate filter. This study also developed and evaluated an IoT-based control system for possible use in solar dryers. An existing solar dryer was modified by introducing the selected adsorbent filter and the developed IoT control system. The performance of the modified solar dryer was evaluated and optimised by determining the effect of the drying kinetics of ginger slices on the quality of the dried products. These were with a view to enhancing the efficiency and effectiveness of the operation of solar dryers. The optimum specification of moisture adsorbent filter was obtained for four adsorbent materials (activated clay, activated charcoal, calcium sulphate, and silica gel). The data obtained for the effect of pack thickness, suction fan speed, and inlet air temperature on the air desiccation performance of the adsorbents was fit in polynomial models and optimised to select the best moisture filter. IoT-based control unit was designed using the Arduino Uno microcontroller, which was interfaced with temperature, humidity, and weight sensors, which were programmed to detect and transmit sensors data to cloud server. The selected absorbent filter as well as the developed control system were then incorporated into the existing mixed-mode solar dryer, and the effect of this modification on the drying kinetics of ginger slices was investigated using response surface methodology. Responses such as total time of active drying and equilibrium moisture content were used as performance indicators of the modified dryer. The optimal conditions for the operation of the dryer, was established and the quality (proximate, phytochemicals and colour) of ginger dried at optimum condition was also determined using standard experimental procedures. The results obtained indicate that silica gel was the most effective adsorbent filter under optimal conditions of 2.03 cm layer thickness, with no requirement for the suction fan. The temperature and relative humidity sensors of the control system were effective, with average accuracies of 98.84% and 96.23%, respectively. However, the weight sensor had an average accuracy of 80.04%. This indicated that the load cell used in the study was sensitive to heat, which adversely affected its accuracy. The performance of the modified solar drying system indicates that the modification significantly aided the drying process of ginger slices, with the best drying conditions being an adsorbent layer thickness of 0.5–1.5 cm and an air velocity of 0.5–2.5 m/s. These conditions resulted in the shortest drying time and a final moisture content of 9.83 to 12.14% wb, which is recommended for safe storage of dried ginger. Nevertheless, the most desirable optimum condition for operating the modified solar dryer was found to be an air velocity of 2.5 m/s and an adsorbent thickness of 1.22 cm, which resulted in a final moisture content of 10.72%. The modification of the dryer significantly influenced the proximate and phytochemical composition of ginger slices. This study concluded that the use of an optimised adsorbent filter and an IoT-based control system can significantly improve the drying process, reduce postharvest losses, and enhance the quality of dried agricultural products.