Development of Mould Core Apple Detection System Using X-ray

Abstract

In order to improve apples quality, many methods have been done on planting, harvesting and storing. However, mould core apples are caused by fungi or/and humid storage condition. Therefore, mould core apples should be selected as the last step to improve the apples quality before selling. This paper describes a detection system depending on X-ray technology, which can get the internal image of the detection objection, to detect the mould core apple. Nevertheless, X-ray is always avoided in the fruit detection because of the radiation of X-ray and the food safety. This paper solves this problem by using a low-level radiation of X-ray radiant. However, the low-level radiation of X-ray radiant generates a low contrast image, which raises a problem in X-ray image processing because it is difficult to distinguish the background and foreground in the low contrast image. In order to perform successful detection of mould core apple, in this paper, firstly, the Wiener filter is utilized for decreasing the image noise from the camera or/and X-ray machine imaging system. Then, the X-ray image is enhanced using contrast enhancement method to solve the problem of low contrast X-ray image. Thirdly, the mould core area is extracted by region recognition and edge detection method after using Lifting Wavelet Transform (LWT). Finally, support vector machine (SVM) is utilized to classify the mould core apple. The experiment results show that the image processing and classification method could detect mould core apple with a success rate of 95%.