Development of a multi-spectral vision system for the detection of defects on apples

Abstract

A method to sort ÔJonagoldÕ apples based on the presence of defects was proposed. A multi-spectral vision system including four wavelength bands in the visible/NIR range was developed. Multi-spectral images of sound and defective fruits were acquired tending to cover the whole colour variability of this bicolour apple variety. Defects were grouped into four categories: slight defects, more serious defects, defects leading to the rejection of the fruit and recent bruises. Stem-ends/calyxes were detected using a correlation pattern matching algorithm. The efficiency of this method depended on the orientation of the stem-end/calyx according to the optical axis of the camera. Defect segmentation consisted in a pixel classification procedure based on the BayesÕ theorem and non-parametric models of the sound and defective tissue. Fruit classification tests were performed in order to evaluate the efficiency of the proposed method. No error was made on rejected fruits and high classification rates were reached for apples presenting serious defects and recent bruises. Fruits with slight defects presented a more important misclassification rate but those errors fitted however the quality tolerances of the European standard. Considering an actual ratio of sound fruits of 90%, less than 2% of defective fruits were classified into the sound ones.

A Survey on Apple Fruit Diseases Detection and Classification

“…They demonstrate that CCV's can give better results than color histograms for image retrieval. [11].In this paper a strategy to sort Jonagold apples relay on the presence of defects was proposed. A multi-spectral vision system incorporating four wavelength bands in the visible/NIR range was created.…”

A sorting optimization curve with quality and yield requirements

“…An example is the automatic identification of contaminants in recycled glass materials (Kattentidt et al, 2003). The second application ranges from the simplest (sorting by size using sieving methods) to the most challenging problems, such as the detection of recent bruises on apples (Kleynen et al, 2005). Other examples of possible applications include the sorting of corn kernels (Liao et al, 1993), rice kernels (Wan, 2002), wheat kernels (Luo et al, 1999), soybeans (Casady et al, 1992;Wang et al, 2002) and ore (Cutmore et al, 1998).…”

“…However, there is an increased interest for the automation of that task as shown in the works of Leemans and Destain (2004), Throop et al (2005), Kleynen et al (2005), Xing and De Baerdemaeker (2007) and Kavdir and Guyer (2008) among others. On the basis of the quality function given in Section 2.2.1 for apples, a sorting optimization curve has been built (Fig.…”

“…The classifier used is based on the Bayes theorem with parametric estimation of probability densities (discriminant analysis). The whole methodology is described by Kleynen et al (2005). Some data from the authors were used to build the SOC The available amount of data not previously used for model training (73) was sufficient to build a SOC, but not as accurately as suggested below.…”

Artificial neural network-based segmentation and apple grading by machine vision

“…Image acquisition and database collection of this work are done in Mechanics and Construction Department of Gembloux Agricultural University of Belgium, and the related details can be found in the works of Kleynen et al [6], [7]. …”