Most ultra wideband (UWB) target detections are mainly carried out in the radiated near field or far field. However, the success of the detection mainly relies on the distance between the sensor and the target which may require a large measurement space. This research investigates the ability to sense targets in the reactive near field in case that the measurement space is constrained. Food detection in a smart fridge is chosen as the main application and test platform. At present, food can be detected by leveraging radio frequency identification (RFID) technology for intelligent fridge. Despite promising results have been shown, it may cause potential health risk and be costly due to tags being affixed to the food to obtain detailed information. Besides, RFID technology lacks of the ability to know the exact food amount such as the level of drink. There is therefore a need for developing new approaches being self content-aware in a low cost and reliable manner. Due to the nature of low cost, relatively high accuracy and immunity to noise, UWB technology provides the potential to detect food as an alternative to RFID. Egg quantity determination, which is an initial and accessible platform of intelligent fridge will be investigated in this thesis. Egg quantity can be well determined in terms of polarisation information in the far field region. However, the challenges arise by taking practical fridge size into account in which the information of eggs will be known in the reactive near field. New approaches are proposed based on investigation of reflection and coupling coefficient correlation of in fridge sensors. Both simulations and measurements are conducted to study the feasibility of sensing the number of egg in the free space environment. Further to this, the effect of other food placed around and above the egg box is investigated in order to verify the robustness of the proposed approaches. Finally, the study is extended to examine the capability of determination of liquid volume. In which, S-parameters are measured related to a variety type of drink in their unique topology and liquid level. The correlation coefficients are evaluated and analysed in both magnitude and phase domain exploiting the amount of liquid information that will be of great significant in the development of future smart fridge.