This study investigated the transformations of lemon flavedo tissues during freeze-drying based on high-resolution X-ray micro-computed tomography (μ-CT) and confocal laser scanning microscopy (CLSM). Contrast-enhanced segmentation, traversal statistic algorithmic analysis, and 3D visualization of lemon peel texture are studied. To avoid pseudo-imaging caused by precipitation of contrast enhancer, low concentration cesium iodide was used to improve X-ray contrast and decrease enlarged inter-cellular region. X-ray μ-CT showed the precipitation of contrast enhancers at the intercellular boundary after freeze-drying. The grayscale thresholding method was presented as the most suitable segmentation method for detecting cell walls due to the high noise level caused by low contrast enhancer concentration in the reconstructed images. The faster water loss at the primary drying stage caused the disintegration of oil glands and induced high porosity of 0.7 % in fresh to 0.84 % in 0.5 h freeze-dried. The results suggest that noise filtering based on voxel statistics provides an efficient approach for the inter and intra-cellular microstructural analysis of excessively watery plant tissues with low contrast enhancer concentration.