Almost half a century has passed since the early innovative work on polymer flooding. In the meantime, to increase the recovery factor and as an attempt to better understand the transport phenomena in porous media, reservoir engineers started to combine various mathematical tools with traditional experimental investigations in the late eighties and early nineties. They began with geometry and statistics and later with topology. Only recently, have we seen studies that use mainly two of these aspects. Nonetheless, a combined study, gathering all four aforementioned subjects (polymer flooding, geometrical, topological and statistical analyses) was missing and this thesis is the pioneering effort on that. The main objective of this thesis is to conduct a comprehensive analysis of polymer flooding, owing to the observation that recovery factor alone, as the main traditional criterion to evaluate polymer flooding, does not provide consistent results in microfluidics. To achieve a complete view, the evaluation consists of three categories of statistical, topological and geometrical analysis. These categories cover different aspects of the flooding including its stability, efficiency and phases mobility. Regarding the methodology, sequences of polymer flooding and water flooding were conducted mainly in two rates of 0.01(High rate/HR) and 0.0019(Low rate/LR) [ml/hr] as well as two phases of secondary and tertiary flooding in a uniform micromodel to control the effect of pore network geometry. All these experiments were performed with crude oil of the target field from Vienna basin, which has a high viscosity and TAN number. Flooding records (duration, injection rate and polymer concentration) as well as images of the micromodel were collected as the primary experimental data and later processed to extract more data. The extraction process was mainly done by image processing and image analysis that proved to be extremely useful and necessary. As for the findings, a new quantitative description of the observed flow regimes in secondary flooding was developed by plotting the RF versus fractal number. The fractal number was extracted using morphological image analysis and it corresponds to the complexity of the perimeter of the areal sweep. The author observed that above a critical polymer concentration the stabilizing effects of polymer agent overcomes the instability caused by viscous or capillary fingering. Furthermore, in the tertiary flooding, ganglion dynamic flow regime has been observed and characterized by employing Euler characteristic. It has been noted that oil clusters become immobile around a higher oil saturation for lower injection rate. As the result of this work, a better insight into the displacement patterns of the flooding phenomena has been achieved. Moreover, these findings should and could be used to combine with traditionally modelling techniques of fluid flow to increase the accuracy of the model simultaneously with the reduction of the computational expense. For example, by running the model over a representative elementary volume.