Kiwifruit canker caused by Pseudomonas syringae pv. actinidiae (Psa) is devastating to the kiwifruit industry. Because of the complex pathogenesis of kiwifruit canker, there is still a lack of effective control measures. The aim of this study was to characterize the temporal and spatial changes of biopolymers in the petiole cell walls of healthy and cankered kiwifruit using confocal Raman microscopy (CRM) techniques combined with Chemometrics. We proposed a general investigation of the development of kiwifruit canker and chemical distribution imaging was successfully applied to the analysis of disease dynamics. First, Raman spectra (RS) of healthy and diseased kiwifruit petiole cell walls were acquired at a micron-scale spatial resolution, followed by label-free in situ imaging of the cell wall. The results of RS and chemical imaging showed that the changes of cellulose and lignin on the petiole cell wall of susceptible kiwifruit under Psa stress were time-dependent, and the degradation rate of cellulose and lignin reached about 25% after eight days of incubation. The results of principal component analysis (PCA) and support vector machine (SVM) based on RS showed that healthy, early-stage disease and late-stage disease samples exhibited significant clustering effects and the SVM classification accuracy was as high as 100%. The multivariate curve resolved–alternating least squares (MCR–ALS) is used to optimize the Raman imaging in order to eliminate the noise and interference caused by the peak clamping method and provide a meaningful chemical profile. In conclusion, CRM combined with chemical classification algorithms and imaging analysis has certain feasibility and application prospects in the early warning and cellular-level dynamic characterization of kiwifruit canker.