|Chao, Kuanglin - Kevin Chao|
|Lo, Y Martin|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 4/7/2008
Publication Date: 5/4/2008
Citation: Lefcourt, A.M., Kim, M.S., Narayanan, P., Reese, D., Tasch, U., Rostamian, R., Chao, K., Lo, Y. 2008. Novel Technologies for Processing Apples [abstract]. United Fresh Produce Assocation Meeting, May 4-8, 2008, Las Vegas, Nevada. Interpretive Summary:
Technical Abstract: The desire to develop methods to use machine vision to simultaneous inspect apples for quality issues and for contamination has resulted in the development of a number of new technologies for processing apples. First, a number of imaging techniques and detection methods were developed under laboratory conditions. For commercial application of these technologies, one dilemma is the difficulty of differentiating the visual complex stem and calyx regions from true problem areas. A classic solution to this dilemma is to orient apples prior to imaging so that the locations of the stem and calyx regions in images are known. A number of technologies to orient apples have been developed, however none has proved commercially viable due to error rate, mechanical complexity, cost, or some combination thereof. We discovered that apples could be oriented using their inertial properties by simply rolling the apples down a track consisting of two parallel rails, as rotational velocity increases apples move to an orientation where the stem/calyx axis is perpendicular to the direction of travel and parallel to the plane of the track. Contemplation of methods to incorporate tracks into existing processing lines resulted in the realization that it would be more cost effective to replace existing mechanical transport systems with a system based on the use of the much simpler and cheaper rails. The functions that need to be added to turn the rails into complete systems include loading, imaging, and sorting. The need to dissipate the kinetic energy accumulated by rolling apples suggests the use of a water medium for loading apples and for transporting sorted apples. Apples are commonly unloaded from crates by submersing the crates. Sorting could be accomplished by using air pulses to blow apples off the tracks and into selected water channels; one channel could allow apples where orientation failed or imagine results were ambiguous to be recycled back to the loading area. To image the apples, a system of concave parabolic mirrors was developed to allow images representative of essentially 100% of an apple's surface to be acquired as apples rolled down the track. Mirrors are needed primarily for contamination detection to enable full visualization inside the concave stem and calyx regions. An imaging system was developed that allows for simultaneous acquisition of both reflectance and fluorescence images using a single camera; reflectance images allow detection of damaged apples and fluorescence images allow detection of fecal contamination. Together, these technologies provide all the pieces necessary to develop a novel system for processing apples that is less expensive and provides a wider range of potential sorting criteria compared to existing mechanical sorting systems.