The liquid-phase diffusion coefficient and viscosity are important physical parameters for characterizing mass transfer and transport behavior in liquid systems, both generally depending on solute concentration. Conventional methods usually require multiple experiments or contact-based measurements, making it difficult to continuously obtain concentration-dependent transport parameters over a wide concentration range. In this work, a non-contact and visual optical method is proposed for determining concentration-dependent diffusion coefficients and associated viscosity over a wide concentration range. A liquid-core cylindrical lens is simultaneously used as both a diffusion cell and an imaging element. By analyzing the transient diffusion images of a binary solution, the relationship among image width, refractive index, and concentration is established, and the spatial and temporal concentration profiles are obtained from the same continuous diffusion process. Based on this, the concentration-dependent diffusion coefficient is expressed as a polynomial, D(C) = D_0(1+ \alpha C+ \beta C^2 + \gamma C^3) , where D₀ is the diffusion coefficient at infinite dilution and α, β, and γ are fitting coefficients. The finite difference method is applied to numerically solve Fick’s diffusion equation, and the calculated concentration profiles are compared with experimental profiles. The coefficients corresponding to the minimum concentration deviation are selected to determine D(C). Under the approximate applicability of the Stokes-Einstein relation, the concentration-dependent associated viscosity η(C) is further calculated from the obtained D(C). Experiments are carried out for the glycine-water system over a concentration range of 0–3.0 mol/L at 298.15 and 303.15 K. At 298.15 K, the measured diffusion coefficient is D(C) = 1.071 \times 10^-5 (1- 0.083C+ 0.010C^2) cm²⋅s^–1, and the associated viscosity increased from 0.8870 to 1.0547 mPa⋅s with increasing concentration. At 303.15 K, the measured diffusion coefficient is D(C) = 1.167 \times 10^-5 (1 - 0.068C+0.006C^2) cm²⋅s^–1. The obtained diffusion coefficients and associated viscosity results are in close agreement with reported values. Furthermore, the measured D(C) is used to calculate the spatial and temporal refractive-index profiles, and ray-tracing simulations are performed to reproduce the dynamic diffusion images. The simulated images closely match the experimental images in terms of the sharp imaging position, overall contour, and image-width profiles along the diffusion direction, further demonstrating the reliability of the proposed method. The results indicate that this method provides an efficient and visual approach for characterizing concentration-dependent transport parameters of liquid systems with high accuracy and stability.