Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 22, 2002
Publication Date: N/A
Interpretive Summary: Fresh apples are traditionally sorted manually to remove defective or decayed products based on visual inspection. This process is laborious and often subjective to human error. To improve sorting efficiency and minimize human errors, an automatic online sorting system is highly needed by the apple industry. Although machine vision technology has showed a great promise in various online sorting applications in food and agriculture, sorting defective apples using this technology has been unsuccessful primarily due to the presence of apple stem-end/calyx, which can be falsely recognized as apple defects. Therefore, it is necessary to develop a technology that distinguishes the true defect from the stem-end/calyx. This research paper describes the development of a new technology that enables the machine vision system to distinguish the true defect from the stem-end /calyx of an apple. By using a dual-wavelength (near-infrared and mid-infrared) vision system, images of apple defects and stem-end/calyx were accurately recorded and analyzed. With the special algorithms developed, the effect of apple stem-end/calyx was eliminated, resulting in a significant improvement in the accuracy of defect sorting without the interference of apple stem-end/calyx. Since defect sorting is an important postharvest process, the application of this technology will improve the efficiency of postharvest handling of apple industry and ensure the quality of apple products being supplied to consumers. In addition to apples, the methodology and algorithms developed in this research are also applicable to defect sorting and online inspection of other fruits and vegetables. Consumers of fresh apples and other fruits and vegetables, as well as the fresh produce processors, packers, wholesalers and retailers will benefit from this research.
A near-infrared (NIR) and mid-infrared (MIR) dual-camera imaging approach for on-line apple stem-end/calyx detection is presented in this paper. How to distinguish the stem-end/calyx from the true defect is a persistent problem in apple defect sorting systems. In a single camera NIR imaging approach, the stem-end/calyx of an apple is usually confused with true defects and is often mistakenly sorted. In order to solve this problem, a dual-camera NIR/MIR imaging method was developed. Based on experimental results, the MIR camera can identify only the stem-end/calyx parts of the fruit, while the NIR camera can identify both the stem-end/calyx parts and the true defects on the apple. A fast algorithm is developed to process the NIR and MIR images and then to achieve reliable defect detection for on-line apple sorting. Test results of this method are also presented.