Volume 5, Issue 2, February – 2020 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Digital Image Processing for Recognition of

Imperfections in the Submerged Hull
Antonio Eduardo Assis Amorim Maria Julia Polastri Cacchi
Department of Naval Shipbuiding, Department of Naval Shipbuilding,
College of Technology of Jahu, CEETEPS College of Technology of Jahu, CEETEPS
Jahu,SP, Brazil Jahu, SP, Brazil

Abstract:- Visual inspection in ships is an important In ship docking operations, it is important to verify the
task to identify its integrity and this task can be done process of alignment of the ship with the riding stables,
with underwater digital cameras. Linear filtering requiring visual monitoring, which can be assisted with
techniques can be used to identify edges. This paper digital image processing applications [2]. Another important
shows an implementation of a code to process digital operation that can be used with image processing is the
image using Sobel filter, adjusted by two parameters evaluation of installation and degradation conditions of
and the results show the code highlights points of existing subsea and submerged lines [3]. Some operations in
corrosion or deformation being an important support the field of petroleum engineering are done using a ROV or
tool for the inspection of the ship's structure conditions. divers being monitored visually. Supporting the operation
with digital image processing can generate useful
Keywords:- Image Processing; Filter; Convolution; Sobel information for the operator, reducing staff uptime.
Filter; Edge Detection.
In view of these facts important to naval industry, it is
I. INTRODUCTION relevant to develop a support code for these visual
inspection activities that will serve as a support tool for
Constant ship operation in harsh climatic environments decision making. Such an application must process images
can cause structural damage or corrosion and require in real time, generating operator-relevant information that
periodical inspection. Ship sidewalls suffer from corrosion- could not be extracted into the original images due to low or
erosion effects when sea waves remove small parts of the excess light, turbidity or noise that generate blurred images.
structure as well as corrosion-cavitation when air bubbles
produce a similar effect. Usually linear filters are used that detect edges. The
filtering technique consists of adjusting the pixel values of
Visual inspection operations on ships are complicated the image as a function of the values of the surrounding
as they involve diving or vessel removal activities in dry pixels so that relevant information can be extracted. The
dikes. To reduce human risks, activities can be done using way to adjust values involves the concept of convolution.
underwater remotely operated vehicle – ROV that carry Usually, to reduce the computational effort, in general, color
embedded cameras. Thus, when the vessel is taken out of images are transformed into shades of gray, which express
service for repair, it is important to check the points of the intensity or level of gray[4].
greatest corrosion attack and this is usually done visually.
However, due to water aspects, visual inspection may not There are several filtering techniques applied to
work so well. various fields producing good results. The most common
filters are Laplacian, Prewitt, Sobel and Gaussian [4–6].
Image processing is a technique that allows
manipulating a digital image through operations, improving Therefore, this article deals with the process of
or extracting useful information that can be used for a application development based on Sobel filter, in order to
particular purpose [1]. Image processing can be a tool that generate images that allow extracting relevant visual
can assist the inspector in conducting the analysis. information that attest to the deterioration state of the
vessel's side structure.
In this way, image processing can be applied to the
naval industry for the inspection of vessels, in matters To this end, a picture will be taken of a old pusher and
related to welding, corrosion and structural aspects as well these images will be digitally processed and a new image
as activity of docking or installation of submerged structures will be generated in order to highlight points of corrosion or
in oil wells. deformation of the vessel's side structure.

Another type of occurrence on ships involves minor The application is made in Octave software using the
collisions that can deform the structure of the ship's side and image processing package.
often, visually, go unnoticed. The digital processing
technique can highlight interesting details useful for It is considered that the linear filters can be adjusted to
evaluation. highlight or accentuate points of relevance in the structure

IJISRT20FEB620 www.ijisrt.com 765

Volume 5, Issue 2, February – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
and allow the operator to extract relevant information and to II. STRUCTURE OF THE CODE
assist in decision making.
The code involves the following steps: image
It is hoped that with this product, companies engaged acquisition, application of the Sobel filter and median filter.
in prospecting, inspection, construction and maintenance of Two parameters are applied to the Sobel filter components
vessels will be able to use it as a decision support tool in the and added to the median filter. The figure below shows the
naval field. structure of the code.

Fig 1:- Code Structure.

III. RESULTS

To assess the quality of the code, the side of an out-of-service pusher was analyzed. Its somewhat precarious state allows the
performance of the code to be evaluated. The figure below shows a partial view of the pusher condition and the processed image.
The original image is on the left side and the processed image is on the right.

Fig 2:- The original image (left) and the processed image (right).

Note that the code highlights the points at which corrosion or deformation stands out on the ship's plate.

Fig 3 shows another image. On the left side there is the original image, which does not show much detail. On the right side the
processed image shows accentuated details of the plate's condition. The code works properly.

IJISRT20FEB620 www.ijisrt.com 766

Volume 5, Issue 2, February – 2020 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 3:- Details of plate deformation appear accentuated in the processed image.

In the following figure there is another detail of the pusher. It is observed that the weld details are highlighted in the
processing, as well as the deformation of the plate.

Fig 4:- Details of weld and deformation of vessel side

IV. CONCLUSION REFERENCES

This paper aims to develop and qualitatively evaluate [1]. Tanimoto SA. An interdisciplinary introduction to
the performance of an application focused on digital image image processing: pixels, numbers, and programs.
processing that aims to extract relevant information in the London - England: The MIT Press, 2013. Epub ahead
structure of the ship's side plate, such as corrosion, of print 2013. DOI: 10.5860/choice.50-2706.
deformation and weld details. There are several applications [2]. Rodrigues JM dos SS. Simulador Interactivo de
in the literature that perform digital image processing using Docagem de Navios em Ambiente Tridimensional.
filters. The application developed in this work uses the Instituto Técnico de Lisboa, 2008.
median filter and the Sobel filter. The application is [3]. Boszczowski RB, Piazza F, Munaro M, et al.
promising, showing the deformations present in the plates. Avaliação de Sistemas Submarinos de Distribuição de
The images obtained show that the application can be a Energia. In: Anais do II Cinetel, pp. 510–513.
useful tool to assist inspection activities. Future work is [4]. Gonzalez RC, Woods RE, Eddins SL. Digital Image
intended to enhance the application to highlight corrosion Processing Using Matlab. Gatesmark Publishing. Epub
points or deformations relevant to inspection. The next step ahead of print 2004. DOI: 10.1117/1.3115362.
is to develop a code in which to identify, by means of [5]. Marques O. Practical image and video processing
colors, the regions attacked by corrosion. using matlab. Wiley, 2011.
[6]. Owotogbe JS, Ibiyemi TS, Adu BA. Edge Detection
Techniques on Digital Images - A Review. Int J Innov
Sci Res Technol 2019; 4: 329–332.

IJISRT20FEB620 www.ijisrt.com 767