A novel image-analysis based approach to evaluate some physicochemical and cooking properties of rice kernels

• Image-based results can be used to evaluate implicit properties of rice kernels.
• Correlations between appearances and implicit properties were proposed.
• Kernel images could be used to indicate appropriate cooking/soaking time of rice.
• Degree of parboiling can roughly be determined from image-based results.

Prior to being edible, paddy needs to undergo several processing steps, which inevitably affect the kernel quality, both in terms of the physical, physicochemical and cooking characteristics. Unfortunately, many of the kernel quality parameters are time-consuming and not straightforward to evaluate, rendering the control of the paddy processing quite ineffective. An alternative means, which allows rapid determination of various properties of the kernels are therefore desirable. This study represents the first attempt to correlate image-analysis based apparent characteristics of rice kernels to physicochemical as well as cooking properties of the kernels for effective monitoring of rice processing and cooking processes. Parboiled rice production process was selected as a test process. Image-analysis based results, in terms of the changes of the kernel dimensions (i.e., elongation ratio, width ratio, and swelling) and translucency as well as the fractal dimension of the kernel images, were noted to adequately correlate with the tested physicochemical (i.e., amylose content, water uptake, degree of parboiling, which is closely related to the degree of starch gelatinization) as well as cooking (i.e., texture and cooked level of rice kernels) properties of the rice kernels of two varieties of Thai rice. This represents an attractive evaluation technique that can be adopted by an industry for various properties measurement and quality control.