计算机视觉在芒果品质评价中的应用

Application of computer vision in the quality evaluation for mango

果类的品质一般用果的酸度,可溶性固形物及硬度等指标来评价。但是从某种意义上说水果的颜色也可以间接表示水果品质。在测定芒果成熟期3种指标同时,拍摄数字像片,以计算机图像分析技术为手段,采集芒果颜色参数。研究3种颜色模型中与芒果品质变化相关的颜色特征量,利用统计软件分析,得出合适的颜色参数与芒果品质指标值间关系式,籍此式可由颜色变化推导芒果的品质变化。结果表明所得关系式有效。

As the traditional sense, some maturity indicators of mango were used to describe the maturity degree, such as general fruit acidity, soluble solids contents, and hardness. But mango color can also express the fruit maturity indirectly in a certain degree. The experiment uses the computer image analysis technology as the detection means. It concluded a formula between color parameters and quality indicators after analyzing the color change in three types of color model. The formula can predict the quality change while input the color parameters.