Development of image processing method for evaluation of tempura

面糊的脆度是天妇罗烹饪的重要因素，与口感密切相关。在本研究中，为了分析天妇罗的烹饪过程，我们估计的时间序列变化的过程中，使用图像纹理分析的油表面的闪烁状态。油表面的图像包括在烹饪期间由气泡和波浪引起的闪烁。为了评价这种纹理图像，我们引入了共生矩阵。为了分析水分蒸发与共生现象之间的对应关系，在油温条件下通过实验验证了该方法的有效性，并对蒸煮后的天妇罗面糊进行了图像处理、破裂分析和感官评价，并进行了PLS回归分析。从这些结果中，证实了天妇罗面糊的图像特征与断裂分析之间的高度相关性。

The crisp of batter is an important factor in tempura cooking, and it is closely related to the taste. In this research, in order to analyze the process of tempura cooking, we estimated the process of temporal sequence changes of the shimmering state of the oil surface by using image texture analysis. The image of the oil surface includes shimmering caused by bubbles and waves during cooking. In order to evaluate this texture image, we introduced the co-occurrence matrix. To analyze the correspondence between evaporation of moisture and the co-occurrence by PLS regression, the effectiveness of our proposed method was confirmed by experiments under the condition of the oil temperature.The tempura batter after cooking is evaluated by the image processing, the rupture analysis and the sensory test, and these are also analyzed by PLS regression. From these results, a high correlation between the image feature of the tempura batter and the rapture analysis was confirmed.