The increasing demand for batteries for mobile and stationary energy storage applications require further research to increase the energy density and the lifetime of battery cells. The functionality, performance characteristics and quality of the cell are correlated to the microstructural properties of the battery. There are different defects that can occur inside a Li-ion battery. On the macroscopic scale, the positioning of the jelly roll or stack inside the battery housing and the correct connection of the current collectors and poles are from major importance. Wrong positioning of the jelly roll / stack or welding beads can lead to internal shortcuts and thermal runaways. On the jelly roll level the correct and accurate stacking or winding of the different electrodes influence the quality of the cell and can diminish the cell performance due to delaminations and unnecessarily increase the production costs due to decreased material utilization. On the microscopic level of the electrodes defects like foreign particles, thickness deviations of the electrode coating, pores and cracks inside the coating layer can occur which lead to capacity fading during the batteries lifetime. Foreign particles can be responsible for internal shortcuts and thermal runaways and are a potential safety risk introduced during cell-production. Electro-chemical methods are most commonly used for battery characterisation and for determination of cell performance parameters. However, these methods do not deliver any spatial information on the cell microstructure and on distribution of different defects. Therefore, additional inspection techniques are necessary to get a deeper knowledge of the interaction of battery microstructure and cell performance and to identify the most important microscopic features determining the cell quality. We show how a combination of different imaging inspection methods can help to get a deeper understanding of cell quality and performance. For an overview of whole battery cells and the detection of large geometric deviations in the battery cell setup non-destructive computer tomography (CT) methods are used. For further investigations on stack or electrode level, we use light microscopy and secondary electron images. high-resolution images with light microscopy of whole cross sections of round cells (18650) or large prismatic cells for plug-in hybrid vehicles (PHEV) are acquired. To process the large amount of image data trainable machine learning algorithms for automated defect detection are applied. With the help of automated defect detection, we can generate important information on different failures and their distribution in Li-ion batteries, which are not accessable with common inspection methods. The used methods deliver important information of the quality and stability of electrode and cell production processes and can help to reduce scrap rates during cell production or can help to evaluate different battery suppliers.