Wheat is the primary grain crop in our country and plays a pivotal role in developing the national economy. However, abiotic stress factors such as salt and physical damage have gradually become essential factors restricting wheat yield and quality. Studies have shown that the cell wall is an important line of defence for plant cells to resist adversity and stress directly. Under salt stress, cell osmotic pressure increases, and the permeability of the plasma membrane will be affected to a certain extent. In order to maintain cell morphology and structure, pectin and other polysaccharides in plant cell walls will be transformed and changed to varying degrees. Physical damage will deepen the degree of lipid peroxidation of plant cell membranes, increase membrane permeability, and lead to the loss and degradation of nutrients. The damaged site and its surrounding cells will also be embolized to block the invasion of bacteria. Pectin, which is the main component of cell wall and can reflect the integrity and permeability of the plant cell wall and membrane system, can be used as an important factor in studying the response mechanism of plant internal substances under stress. At present, commonly used pectin detection methods such as Gravimetric Colorimetric, Liquid chromatography, etc. are cumbersome to operate, not real-time, and large sample consume. There is an urgent need for a simple, fast and non-destructive detection method. This paper used hydroponics to research the wheat (Yannong 0428). The sodium chloride (NaCl) solution was applied to culture medium, and acupuncture was carried out on both sides of the main vein of the first leaf of wheat to simulate the salt damage and the physical damage caused by insect bites, respectively. The pectin and its hyperspectral information of wheat leaf were also collected and processed. The correlation analysis method was used to screen the sensitive spectral band. The three modeling methods of principal component regression (PCR), partial least squares (PLS), and stepwise multiple linear regression (SMLR) were combined with multiple scattering correction processing (MSC), standard normal transformation processing (SNV), first derivative processing (FD), convolution smoothing (SG), and Norris derivative filter processing (NDF) to establish a pectin content inversion model. Finally, the model established by PLS+SNV+FD+NDF method was selected as the optimal model, and its performance was also tested. The results showed that the predicted value of pectin content was consistent with the measured value, and the coefficient of determination (R²) and root mean square error (RMSE) were determined to be 0.997 6 and 0.35, respectively. The repeatability of the predicted value was good, and the relative standard deviation (RSD) was 1.2%. This study uses a new method to realize the high-precision, fast and non-destructive detection of wheat pectin, which is helpful to the in-depth study of the mechanism of the wheat response to stress and provides a reference for the prediction of the stress degree of field crops and the accurate control of planting environment.