Use of Digital Image Analysis as a Monitoring Tool for Non-Uniform Deformation of Shrinkable Materials during Drying

Development and use of three-dimensional image analysis algorithms to evaluate puffing of banana slices undergone combined hot air and microwave drying

“…Two-dimensional images of banana slices were first taken via the use of a scanner (Epson, V30, Indonesia) at 300 dpi with black background to reduce the shadow. Threedimensional images were produced from the two-dimensional images as per the procedures developed by Jinorose et al [3] Two-dimensional images were preprocessed by MeshLab (ISTO-CNR, Visual computing Laboratory) to reduce unwanted objects and then reconstructed using Autodesk Recap 360 software (Autodesk Inc., San Rafael, CA) into three-dimensional images.…”

“…[2] However, volumetric deformation cannot be used to describe non-uniform or irregular puffing, which normally takes place and can significantly affect the appearance and hence the consumer's acceptance of a final product. [3] It has in fact been reported that two pieces of a material may exhibit similar volumetric shrinkage (or deformation) despite the fact that they had gone through different methods of drying and clearly possess different forms (shape and size) of deformation. [4] Although image analysis has been used in some earlier studies to evaluate deformation during drying, attempts were usually made only to evaluate deformation in one or two dimensions, which is not adequate when a sample deforms non-uniformly [3] or exhibits irregularly rugged surface in three dimensions such as in the case of puffing.…”

“…[3] It has in fact been reported that two pieces of a material may exhibit similar volumetric shrinkage (or deformation) despite the fact that they had gone through different methods of drying and clearly possess different forms (shape and size) of deformation. [4] Although image analysis has been used in some earlier studies to evaluate deformation during drying, attempts were usually made only to evaluate deformation in one or two dimensions, which is not adequate when a sample deforms non-uniformly [3] or exhibits irregularly rugged surface in three dimensions such as in the case of puffing. This is simply because one and two dimensional imagings are much easier than three-dimensional imaging, eventhough they cannot accurately well represent deformation.…”

“…This is simply because one and two dimensional imagings are much easier than three-dimensional imaging, eventhough they cannot accurately well represent deformation. [3] In addition, despite some recent attempts to describe deformation using indicators derived from imagebased information, most studies only focused on uniform deformation of simple shapes, e.g., spherical, cylindrical and cubical shape. [3] The ability to precisely identify the shape and its changes of an irregularly shaped materials remains a challenge.…”

Development and testing of a novel image analysis algorithm for descriptive evaluation of shape change of a shrinkable soft material

“…Image analysis is an alternative method that can be used for material and product characterization; the method can well be used for the determination of both size and shape of a product 14 , 15 . Nevertheless, although there exist several algorithms that can be used to identify shape, e.g., template matching method 14 , or such statistical methods as k-cluster and discriminant analyses 14 , 16 as well as decision tree 14 , 17 – 19 , such available algorithms can only classify images into a limited number of predetermined shapes, making these algorithms less robust and flexible.…”