The focus of this thesis is the improvement of acquisition and processing of Mueller matrix polarimetry using a confocal scanning laser ophthalmoscope (CSLO) and the application of Mueller matrix polarimetry to image the retina. Stepper motors were incorporated into a CSLO to semi-automate Mueller matrix polarimetry and were used in retinal image acquisition. Success rates of Fourier transform based edge detection filters, designed to improve the registration of retinal images, were compared. The acquired polarimetry images were used to reassess 2 image quality enhancement techniques, Mueller matrix reconstruction (MMR) and Stokes vector reconstruction (SVR), focusing on the role of auto-contrasting or normalization within the techniques and the degree to which auto-contrasting or normalization is responsible for image quality improvement of the resulting images. Mueller matrix polarimetry was also applied to find the retardance image of a malaria infected retinal blood vessel imaged in a confocal scanning laser microscope (CSLM) to visualize hemozoin within the vessel. Image quality enhancement techniques were also applied and image quality improvement was quantified for this blood vessel. The semi-automation of Mueller matrix polarimetry yielded a significant reduction in experimental acquisition time (80%) and a non-significant reduction in registration time (44%). A larger sample size would give higher power and this result might become significant. The reduction in registration time was most likely due to less movement of the eye, particularly in terms of decreased rotation seen between registered images. Fourier transform edge detection methods increased the success rate of registration from 73.9% to 92.3%. Assessment of the 2 MMR images (max entropy and max signal-to-noise ratio (SNR)) showed that comparison to the best CSLO images (not auto-contrasted) yielded significant average image quality improvements of 158% and 4% when quantified with entropy and SNR, respectively. When compared to best auto-contrasted CSLO images, significant image quality improvements were 11% and 5% for entropy and SNR, respectively. Images constructed from auto-contrasted input images were of significantly higher quality than images reconstructed from original images. Of the 2 other images assessed (modified degree of polarization (DOPM) and the first element of the Stokes vector (S0)), DOPM and S0 yielded significant average image quality improvements quantified by entropy except for the DOPM image of the RNFL. SNR was not improved significantly when either SVR image was compared to the best CSLO images. Compared to the best auto-contrasted CSLO images, neither DOPM nor S0 improved average image quality significantly. This result might change with a larger number of participants. When MMR were applied to images of malaria infected retinal slides, image quality was improved by 19.7% and 15.3% in terms of entropy and SNR, respectively, when compared to the best CSLO image. The DOPM image yielded image quality improvements of 8.6% and -24.3% and the S0 image gave improvements of 9.5% and 9.4% in entropy and SNR, respectively. Although percent increase in image quality was reduced when images were compared to initial auto-contrasted CSLO images, the final image quality was improved when auto-contrasting occurred prior to polarimetry calculations for max SNR and max entropy images. Quantitative values of retardance could not be found due to physical constraints in the CSLM that did not allow for characterization of its polarization properties and vibrational noise. Mueller matrix polarimetry used to find the retardance image of a malaria infected retina sample did yield visualization of hemozoin within the vessel but only qualitatively. In conclusion, improvements in the acquisition and registration of CSLO images were successful in leading to considerably shorter experimentation and processing times. In terms of polarimetric image quality improvement techniques, when compared to the best CSLO image. A large proportion of the improvement was in fact due to partially or completely stretching the pixel values across the dynamic range of the images within the algorithm of each technique. However, in general the image quality was still improved by the Mueller matrix reconstruction techniques using both entropy and SNR to generate the CSLO retinal images and the CSLM imaged malaria infected sample. In the malaria sample, retinal blood vessel visualization was also qualitatively improved. The images yielded from Mueller matrix polarimetry applied to a malaria infected retinal sample localized hemozoin within the blood vessel, but a quantitative image of the phase retardance could not be achieved.

The focus of this thesis is the improvement of acquisition and processing of Mueller matrix polarimetry using a confocal scanning laser ophthalmoscope (CSLO) and the application of Mueller matrix polarimetry to image the retina. Stepper motors were incorporated into a CSLO to semi-automate Mueller matrix polarimetry and were used in retinal image acquisition. Success rates of Fourier transform based edge detection filters, designed to improve the registration of retinal images, were compared. The acquired polarimetry images were used to reassess 2 image quality enhancement techniques, Mueller matrix reconstruction (MMR) and Stokes vector reconstruction (SVR), focusing on the role of auto-contrasting or normalization within the techniques and the degree to which auto-contrasting or normalization is responsible for image quality improvement of the resulting images. Mueller matrix polarimetry was also applied to find the retardance image of a malaria infected retinal blood vessel imaged in a confocal scanning laser microscope (CSLM) to visualize hemozoin within the vessel. Image quality enhancement techniques were also applied and image quality improvement was quantified for this blood vessel. The semi-automation of Mueller matrix polarimetry yielded a significant reduction in experimental acquisition time (80%) and a non-significant reduction in registration time (44%). A larger sample size would give higher power and this result might become significant. The reduction in registration time was most likely due to less movement of the eye, particularly in terms of decreased rotation seen between registered images. Fourier transform edge detection methods increased the success rate of registration from 73.9% to 92.3%. Assessment of the 2 MMR images (max entropy and max signal-to-noise ratio (SNR)) showed that comparison to the best CSLO images (not auto-contrasted) yielded significant average image quality improvements of 158% and 4% when quantified with entropy and SNR, respectively. When compared to best auto-contrasted CSLO images, significant image quality improvements were 11% and 5% for entropy and SNR, respectively. Images constructed from auto-contrasted input images were of significantly higher quality than images reconstructed from original images. Of the 2 other images assessed (modified degree of polarization (DOPM) and the first element of the Stokes vector (S0)), DOPM and S0 yielded significant average image quality improvements quantified by entropy except for the DOPM image of the RNFL. SNR was not improved significantly when either SVR image was compared to the best CSLO images. Compared to the best auto-contrasted CSLO images, neither DOPM nor S0 improved average image quality significantly. This result might change with a larger number of participants. When MMR were applied to images of malaria infected retinal slides, image quality was improved by 19.7% and 15.3% in terms of entropy and SNR, respectively, when compared to the best CSLO image. The DOPM image yielded image quality improvements of 8.6% and -24.3% and the S0 image gave improvements of 9.5% and 9.4% in entropy and SNR, respectively. Although percent increase in image quality was reduced when images were compared to initial auto-contrasted CSLO images, the final image quality was improved when auto-contrasting occurred prior to polarimetry calculations for max SNR and max entropy images. Quantitative values of retardance could not be found due to physical constraints in the CSLM that did not allow for characterization of its polarization properties and vibrational noise. Mueller matrix polarimetry used to find the retardance image of a malaria infected retina sample did yield visualization of hemozoin within the vessel but only qualitatively. In conclusion, improvements in the acquisition and registration of CSLO images were successful in leading to considerably shorter experimentation and processing times. In terms of polarimetric image quality improvement techniques, when compared to the best CSLO image. A large proportion of the improvement was in fact due to partially or completely stretching the pixel values across the dynamic range of the images within the algorithm of each technique. However, in general the image quality was still improved by the Mueller matrix reconstruction techniques using both entropy and SNR to generate the CSLO retinal images and the CSLM imaged malaria infected sample. In the malaria sample, retinal blood vessel visualization was also qualitatively improved. The images yielded from Mueller matrix polarimetry applied to a malaria infected retinal sample localized hemozoin within the blood vessel, but a quantitative image of the phase retardance could not be achieved.

The problem, to make a study of the petroleum industry in Ector County, Texas, and, in addition, a study of the terminology of this industry, was selected as a result of the realization that the major industry of the community was not understood by the teachers, and that it did not have a place in the curriculum.

An electrical driven organic solid state laser is a very challenging goal which is so far well beyond reach. As a step towards realization, we monolithically implemented an Organic Light Emitting Diode (OLED) into a dielectric, high quality microcavity (MC) consisting of two Distributed Bragg Reflectors (DBR). In order to account for an optimal optical operation, the OLED structure has to be adapted. Furthermore, we aim to excite the device not only electrically but optically as well. Different OLED structures with an emission layer consisting of Alq3:DCM (2 wt%) were investigated. The External Quantum Efficiencies (EQE) of this hybrid structures are in the range of 1-2 %, as expected for this material combination. Including metal layers into a MC is complicated and has a huge impact on the device performance. Using Transfer-Matrix-Algorithm (TMA) simulations, the best positions for the metal electrodes are determined. First, the electroluminescence (EL) of the adjusted OLED structure on top of a DBR is measured under nitrogen atmosphere. The modes showed quality factors of Q = 60. After the deposition of the top DBR, the EL is measured again and the quality factors increased up to Q = 600. Considering the two 25-nm-thick-silver contacts a Q-factor of 600 is very high. The realization of a suitable encapsulation method is important. Two approaches were successfully tested. The first method is based on the substitution of a DBR layer with a layer produced via Atomic Layer Deposition (ALD). The second method uses a 0.15-mm-thick cover glass glued on top of the DBR with a 0.23-µm-thick single-component glue layer. Due to the working encapsulation, it is possible to investigate the sample under ambient conditions.

The potential college students of the Odessa area need to know which type of school best suits their personal needs. The area around Odessa offers the highest paid secretarial Jobs in the state, with some 500 oil and oil supply firms employing from one to fifty full-time secretaries, with salaries ranging from $200 to 4325 per month. The opportunities for accountants, bookkeepers and auditors in the Odessa area is unlimited, and they earn from 4350 to $450 per month. The Odessa area also employs approximately 30 business teachers. 1 The. high school graduate is often at a loss as to whether to take advantage of a business course in a private business college in Odessa or whether to study business administration for one or two years at the local junior college, either in preparation for a job or for additional education at another college or university. A great deal of time and money may be wasted if a potential student attends the school that does not fulfill his needs. In view of the above, the purpose of this study is to make comparisons between the two types of schools in an effort to determine the type of school most suited for each student, and to pave the way for improvements in business education in that area. These facts will be helpful in planning the future curriculum and expansion of Odessa College and the possible establishment of a business school in connection with Odessa College.