Study of dimensional changes in rice kernels during milling, soaking and cooking using image analysis | |
| Author | Yadav, Binod Kumar |
| Call Number | AIT Diss. no.PH-04-01 |
| Subject(s) | Rice--Milling Cookery (Rice) |
| Note | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering. School of Environment, Resources and Development |
| Publisher | Asian Institute of Technology |
| Series Statement | Dissertation ;|vno. PH-04-01 |
| Abstract | Importance of two-dimensional image analysis is now widely recognized for determining the quality attributes of cereals and grains linked with their size, shape and appearance. Production of edible rice requires several processing operations such as milling, soaking and cooking during which significant changes occur in kernel dimensions and appearance. However, little information is available in published literature on the applications of image analysis techniques for monitoring the dimensional changes in rice kernels during processing in relation to the varietal differences manifested by the physicochemical properties. This study was aimed at investigating the changes in the dimensions of rice kernels and appearance by image analysis during milling, soaking and cooking. In milling, the emphasis was on the estimation of head rice yield (HRY) defined as the proportion by weight of milled kernels with three-quarters or more of their original length along with kernel whiteness (degree of milling). Further, the changes in kernel dimensions during soaking and cooking of milled rice were investigated in relation to their physicochemical properties. Finally, interrelationships among dimensional changes, water uptake and physicochemical properties of milled rice kernels during soaking and cooking were developed. An image analysis setup consisting of personal computer, frame grabber and a color CCD camera was used for the measurement of kernel dimensions and gray level distribution in two-dimensional images with ImageTool 2.0 software available in public domain. Ten Thai rice varieties ranging from low to high amylose content (16 to 28% d.b.) were selected for the study. A total of 50 samples, 5 for each variety, of rough rice weighing 250 g each was dehusked using Satake dehusker and milled for 0.5 to 2.5 min at an interval of 0.5 min with Satake polisher for obtaining various levels of head rice yield (HRY) and degree of milling. A representative sample of milled rice comprising head and broken kernels, and weighing about 12 g was used for imaging by placing the kernels under CCD camera manually to prevent the kernels touching from one another. Bulk samples of milled rice, which had been subjected to different degrees of milling, were imaged for determining the gray level distribution. Well-milled whole kernel rice was conditioned at three initial moisture contents (Mo) of about 8, 12 and 16% (d.b.) for soaking and cooking experiments. A total of 105 kernels were imaged at each time interval to monitor the dimensional features such as length, width, perimeter and projected area during soaking whereas a total of 32 kernels were imaged during cooking for each variety at the selected M0• Water uptake by 2 g rice kernels during soaking was determined in terms of change in their moisture content. Physicochemical properties, namely, amylose content (AC), gel consistency (GC), alkali spreading value (AS) and protein content (PC) of the rice varieties were determined by standard methods. The selected rice varieties were long grain type having mean kernel length, width, perimeter and projected area in respective ranges of 6.5-7.1 mm, 1.92-2.13 mm, 15.75- 16.94 mm and 11.18-12.89 mm 2 at Mo of 8% ( d.b ). Comparison of mean length of 100 whole kernels of all varieties showed that they formed six significantly different groups at 95% confidence interval. The standard deviation of length distribution for the rice varieties ranged from 0.22 to 0.35 mm indicating considerable overlapping in the distribution of kernel length. ii ii All dimensional measurements made on whole kernels showed a decrease in their mean values with increase in milling duration according to a power-law relationships. However, the normalized measurements of kernel dimensions did not show significant relationships with their physicochemical properties. It was found that the HRY was a power function of the characteristic dimension ratio (CDR) of milled rice samples defined as the ratio of the sum of dimensional measurements on head rice to that on total milled rice kernels in the sample. The CDR based on kernel projected area estimated the HRY with the highest accuracy (R2 = 0.994, SEE= 1.1%). The degree of milling in terms of whiteness index (WI) of the milled rice sample measured by a commercial whiteness meter was a linear function of mean gra)' level (MOL) extracted from the 8-bit images of bulkmilled rice samples indicating an R2 value of 0.99. During soaking in water at room temperature of about 26°C, the changes in milled rice kernel dimensions took place at a faster rate in the beginning and were followed by a diminishing rate finally leading to an equilibrium state. The percent increase in the dimensional measurements of high amylose rice varieties after attaining equilibrium was apparently lower as compared to the low and intermediate amylose rice varieties. It was observed that kernel shape did not change during entire soaking process and the changes in kernel dimensions with soaking duration could be best expressed by a modified exponential relationship for all rice varieties. The regression coefficients of the fitted relationship for describing changes in kernel dimensions were found . to the function of initial moisture content and physicochemical properties of different rice varieties used in the experiments. Water uptake also followed the trend similar to dimensional measurements during soaking of rice kernels and could be best expressed by an exponential relationship. The equilibrium moisture content of milled rice (Mc) depended upon the initial moisture content (Mo) and amylose content. The parameters of the fitted exponential model could be estimated from the physicochemical properties, namely, AC and PC of different rice varieties. In cooking experiments, the milled rice kernels exhibited two stages of changes in their dimensions most possibly due to the pre- and post-gelatinization of the starch granules resulting in two distinct shape factors. The changes in rice kernel dimensions with cooking time were also expressed by an exponential relationship for different rice varieties. Though the changes in rice kernel dimensions were apparently affected by the physicochemical properties, they could not be quantified adequately. The water uptake by rice kernels during cooking could also be modeled by exponential relationships for all rice varieties. The stepwise regression showed the parameters of fitted models had distinct relationships with the physicochemical properties such as AC and AS. The solid loss in cooking gruel was also correlated with AC and AS of milled rice samples. It was observed that the dimensional parameters of the milled rice kernel exhibited a power-law relationship with their moisture content during cooking. It was concluded that the image analysis could pave a way towards on-line monitoring of HRY and whiteness during milling. The equilibrium water uptake and dimensional changes during soaking could be helpful in screening the rice varieties with high, intermediate and low amylose contents. However, the dimensional changes in rice kernels during cooking did not show significant relationships with their physicochemical properties. The dimensions of milled rice kernels at any time during soaking and cooking could be approximated from their moisture content. Also, the moisture content of milled rice kernels during soaking and cooking could be estimated from their physicochemical properties. |
| Year | 2004 |
| Corresponding Series Added Entry | Asian Institute of Technology.|tDissertation ; no. PH-04-01 |
| Type | Dissertation |
| School | School of Engineering and Technology (SET) |
| Department | Other Field of Studies (No Department) |
| Academic Program/FoS | Postharvest and Food Process Engineering (PH) |
| Chairperson(s) | Jindal, Vinod K. |
| Examination Committee(s) | Athapol Noomhorm ;Honda, Kiyoshi;Gunasekaran, Sundaram |
| Degree | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering. School of Environment, Resources and Development |