Location: Sugarbeet and Bean ResearchTitle: Fast bi-dimensional empirical mode decomposition as an image enhancement technique for fruit defect detection
|LU, YUZHEN - Michigan State University|
Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/16/2018
Publication Date: 7/17/2018
Citation: Lu, Y., Lu, R. 2018. Fast bi-dimensional empirical mode decomposition as an image enhancement technique for fruit defect detection. Computers and Electronics in Agriculture. 152:314-323.
Interpretive Summary: Computer imaging technique is now widely used in commercial packinghouses for sorting and grading fruit based on color, size or shape, and presence or absence of surface and subsurface defect. It is, however, still challenging for accurate detection of surface and subsurface defects for apples and other horticultural products because there exist a large variety of defects with different morphological characteristics and because the acquired images often contain artifacts due to uneven illumination and non-flat, irregular fruit surface. These image artifacts, if not removed, can cause serious classification errors (e.g., normal tissues being recognized as defective, and vice versa). Hence enhancement of the acquired images by removing these artifacts is critical for accurate detection of fruit defect. Currently, removal of image artifacts is usually achieved through better design of the illumination system and use of a mathematical model for correcting the fruit shape effect. While useful, these correction methods are cumbersome and inflexible, and often cannot achieve desired results. In this research, a new method, based on bi-dimensional empirical mode decomposition (BEMD), was proposed for removing artifacts from the images acquired using a newly developed structured illumination reflectance imaging (SIRI) system, which provides additional, more detailed information about the imaged objects, compared to the conventional imaging systems under uniform illumination. Computer simulation and experiment on several varieties of apples with pre-existing surface and subsurface defects were performed to evaluate the proposed image enhancement method, along with three other existing methods. Results showed that the new image enhancement method was faster in the image processing, compared to the three existing methods, while achieving the same level of image enhancement. The method was effective in removing the artifacts caused by the fruit shape. The enhanced images showed more distinctive defect features, which would improve image segmentation of defects from normal tissues on the apples. The method is useful for enhancing SIRI-acquired images for defect detection, and it is also promising for use with other conventional imaging modalities for enhanced fruit defect detection.
Technical Abstract: Image enhancement is critical to detection of fruit defects by imaging techniques. Vignetting and noise are major image artifacts, which can seriously affect image segmentation results, especially in inspecting the curved-surface objects like fruit. It is common to use a calibration object or a mathematic model to reduce the vignetting and noise effect in defect detection, but the approach is often cumbersome, inflexible, and difficult to achieve desired results. In this study, a new image enhancement method based on bi-dimensional empirical mode decomposition (BEMD) of images was proposed to isolate and subsequently remove the effects of vignetting and noise by means of selective image reconstruction. The new BEMD method, along with three other BEMD methods, was first tested on decomposing a synthetic image with artificially added vignetting background and noise. The BEMD was found to be the most efficient for image decomposition in terms of computation time, and also give high-quality reconstructed images. Experiments were further conducted by applying the BEMD to the direct and amplitude component images of apple samples with subsurface bruising and surface defects, which were acquired by using a structured-illumination reflectance imaging (SIRI) system. BEMD effectively reduced the image vignetting and greatly enhanced the defect features of the apples, based on both visual inspection and quantitative evaluation. BEMD offers an effective tool for enhancing SIRI images, and it is also promising for image enhancement with other imaging modalities for fruit defect detection.