INTRODUCTION:

Hookworm infection seriously threatens human health, causing intestinal inflammation, progressive iron/protein-deficiency anemia, mucosa damage, and malnutrition of human. Hookworm infection in pregnancy can cause retarded growth of the fetus premature birth and low birth weight. Hookworm in children can cause intellectual, cognitive and growth problems. As one of the most common human helminths, hookworm is a kind of small tubular structure with greyish white or pinkish semi-transparent body. Henceforth it’s necessary to detect hookworm in order to safeguard the human from dreadful diseases. Automatic hookworm detection is a challenging task due to poor quality of images, presence of extraneous matters, complex structure of gastrointestinal and diverse appearances in terms of Colour and texture.

In this work the automated neural network based recognition of hookworm with image processing has been suggested. This technique widely used diagnostic technique to examine inflammatory bowel diseases and disorders. The various images of hookworm which are infecting the intestine are collected from different internet databases. The image processing techniques has been applied for the collected image. Based on the segmented image the features are extracted. The statistical parameters which are extracted is used to train the neural network which classify the presence of hookworm from the image which has been intimated to the user through the mail. This leads the detection of hookworm in its first stage of investigation in the intestine. The results are validated by simulation as well as through Raspberry pi hardware with camera which in turn captures real time images from the environment. This project helps the patient to diagnosis the disease in its initial stage and support to eradicate the disease.

EXISTING WORK:

HOOKWORM:

It is an intestinal parasite that usually causes diarrhea cramps. Heavy infestation with hookworm can be serious for newborns, children, pregnancy women, and persons who are malnourished. Hookworm infections occur mainly in tropical and subtropical climates and affect about 1 billion people -- about one-fifth of the world's population. Hookworms have a complex life cycle that begins and ends in the small intestine. Hookworm eggs require warm, moist, shaded soil to hatch into larvae. This journey takes about a week. In the small intestine, the larvae develop into half-inch-long worms, attach themselves to the intestinal wall, and suck blood. The adult worms produce thousands of eggs. These eggs are passed in the feces (stool). If the eggs contaminate soil and conditions are right, they will hatch, molt, and develop into infective larvae again after 5 to 10 days.

LITERATURE SURVEY:

Jun-Yan He, et.al (2018), suggested this paper, a novel deep hookworm detection framework (DHDF) is proposed for WCE images, which simultaneously models visual appearances and tubular patterns of hookworms¹. This is the first deep learning framework specifically designed for hookworm detection in WCE images. Two CNN networks, namely edge extraction network and hookworm classification network, are seamlessly integrated in the proposed framework, which avoid the edge feature caching and speed up the classification¹.

Drawbacks:

· It can be suitable for simple statistical model.

· The method is fails to recognize that boundary is smooth.

Author:

Priyanka K, et.al (2018), it is proposed to examine the order of capsule endoscopic images using neural network approaches². Information obtaining for the proposed classifier intended for the Recognition of Capsule endoscopic images. Image information will be collected from the diverse distinctive labs. The most vital unconnected highlights and additionally coefficient from the images will be removed. so as to separate highlights, measurable strategies, image handling systems, changed area will be utilized².

Drawbacks:

· The image preparing innovation can be isolated from the innovation of WCE, thus it won't be presented here.

Author:

S. Haritha, et.al (2017), in this paper it describes that endoscopy constitutes the WCE has marked a revolution in the field of GI imaging [3]; beginning an era of non-invasive is utilization of the GI tract and, especially, the entire small bowel. One of the most common GI medical findings efficiently detected by WCE is an ulcer. Approximately 10% of the people suffer from ulcerations. The main disadvantage of WCE is the time-consuming task of reviewing the 55.000 images produced [3].

Drawbacks:

· Wireless capsule endoscopy (WCE) is a radical, patient-friendly imaging system that aids non-invasive photographic review of the patient’s digestive tract and, especially, small intestine.

· However, reviewing the endoscopic data is time-consuming and requires the intense labour of highly experienced physicians.

Author:

R. Sneha, et.al (2017), in this paper it is examined that Detect the presence of the hookworm in the digestive track of the human being and also highlights the location of the worm [4]. The Wireless capsule Endoscopy images are given as the input. They are coloured images. During the pre-processing the coloured images are converted to gray-scale. As a part of segmentation, the Region of Interest (ROI) is extracted from the gray-scale images. Once the features are extracted, the back propagation neural network is used to classify the hookworm images from the normal images among the dataset. In order to localize the worm, the image enhancement techniques like Adaptive Histogram and Morphological operations like Dilation and Erosion are implemented [4].

Drawbacks:

· In this project FDCT, is used they are low and high frequency curve let’s. These curvelets are also called as the sub-bands.

· The Low frequency curvelets do not provide more information about the image.

· Hence only the high frequency curvelets are used for the next phase.

Author:

R. Sasi, et.al (2017), in this project by observing its unique properties, we propose serials of novel techniques to capture its characteristics, aiming to reduce the number of images a clinician needs to review [5]. Experiments from different aspects demonstrate that the proposed method is a robust classification tool for hookworm detection, which achieves promising performance. The contributions of this work are as follow, Guided filter is used to enhance the image, and Guided filter is non-iterative, fast, accurate edge preserving filtering [5].

Drawbacks:

This project can’t detect various intestinal disorders using various algorithms and filters for clearer view.

Author:

Eva Tuba, et.al (2017), in this paper we use this colour space. In this colour space hue and saturation describe chrominance, while luminance is represented as intensity [6]. Each pixel is described by these three components (hue, saturation intensity). Dominant colour is defined by hue component, while saturation measures colourfulness. HSI colour space is also suitable because it has been shown that it is colour invariant, e.g. the same Euclidian distance correspond to the same perceiving differences of human eyes in any part of the colour space [6].

Drawbacks:

Some other texture features can’t be tested as well as detection of other abnormalities.

Author:

M. Yesodha, et.al (2017), in this paper the given image is in RGB colour which must be converted into gray-scale images [7]. The ROI is extracted, that is the area where the hookworm is present. This image is decomposed into several sub bands using fast discrete curvelet transform. Artificial Neural Networks (ANN) forms an appropriate internal feature extractors and classifiers based on the input images. The classifiers get the GLCM values of the input image and then these images are compared with the images present in the database. The database images consist of several images whose GLCM values are been calculated. By comparing these images the hookworm is detected from it. Adaptive histogram improves local contrast and enhancing definition of edges in each region of the image [7].

Drawbacks:

This method of detection of hookworm is less efficient when more number of images should be analyzed at the same time.

Author:

Saraswathy.N, et.al (2017), in this paper is to deals with the determination of hookworms from the Wireless Capsule Endoscopic (WCE) images [8]. The proposed system for the revelation of hookworm is Graph cut method in segmentation stage and neural network as classifier. Our objective is to detect the hookworm more accurately and to reduce the number of clinician images using Multi-scale Dual Matched Filter (MDMF) in WCE images. This automatic detection system can be used in real conditions to assist endoscopists [8].

Drawbacks:

Tubular detection is done with MDMF that with segment only the tubular structure by eliminating the noise.

Author:

Xiao Wu, et.al (2016), in this paper it describes that Wireless capsule endoscopy (WCE) has become a widely used diagnostic technique to examine inflammatory bowel diseases and disorders [9]. As one of the most common human helminths, hookworm is a kind of small tubular structure with grayish white or pinkish semi-transparent body, which is with a number of 600 million people infections around the world. Automatic hookworm detection is a challenging task due to poor quality of images, presence of extraneous matters, complex structure of gastro-intestinal and diverse appearances in terms of colour and texture. This is the first few works to comprehensively explore the automatic hookworm detection for WCE images. To capture the properties of hookworms, the multi scale dual matched filter is first applied to detect the location of tubular structure. Piecewise parallel region detection method is then proposed to identify the potential regions having hookworm bodies. To discriminate the unique visual features for different components of intestinal, the histogram of average intensity images [9].

Drawbacks:

Good performance has been achieved, there are still 23% of hookworm images cannot be correctly detected.

I. PROPOSED WORK:

DETECTION OF HOOKWORM USING IMAGE PROCESSING:

Fig: 1: Block Diagram

INPUT IMAGE:

The fig.1 represents the block diagram of image acquisition and processing

· Some Hookworm affected patient images are taken as the dataset image for the input to make the processing.

· The dataset input image contains of some normal affected hookworm image as well as abnormal condition of hookworm image.

GRAY IMAGE:

If the Hookworm image is RGB image it must be converted to Grayscale image or black and white image before processing. In RGB image the implementation of colour conversion is difficult.

FEATURES EXTRACTION:

The grayscale images will be subjected to feature extraction that includes size and shape based features also few other statistical parameters

GUIDED FILTER:

The guided filter function performs edge-preserving smoothing on an image, using the content of a second image, called a guidance image, to influence the filtering. The guidance image can be the image itself, a different version of the image, or a completely different image. Guided image filtering is a neighborhood operation, like other filtering operations, but takes into account the statistics of a region in the corresponding spatial neighborhood in the guidance image when calculating the value of the output pixel.

MORPHOLOGICAL OPERATION:

Morphology is a broad set of image processing operations that process images based on shapes. In a morphological operation, each pixel in the image is adjusted based on the value of other pixels in its neighborhood. By choosing the size and shape of the neighborhood, you can construct a morphological operation that is sensitive to specific shapes in the input image. Morphological Dilation and Erosion. The number of pixels added or removed from the objects in an image depends on the size and shape of the structuring element used to process the image.

The guided filter computes the filtering output by considering the content of a guidance image, which can be the input image or another different image. Guided filter uses the colour images for implementation because colour guidance image can better preserves the edges that are not distinguishable in gray-scale Gray Level Co-occurrence Matrix (GLCM) is proposed to Classify the hookworm images, whether hookworm is present or not in WCE images. The Morphological operation is used to the segment the Hookworm.

RESULTS AND DISCUSSION:

Selection of input image for the database images. There 100 database images where 60 images are hookworm images, 20 images of other worms inside the body and 20 no worm images. The collected database images will be in the form of RGB images. Colour conversion for RGB image to grayscale image. Guided filter is used for edge extraction. Adaptive histogram equalization is obtained for enhancing the definition of edges in each region of the image. The morphological operation is used for defining the structure of the worm and segmentation of the surface of worm. Morphology is a technique of image processing based on shapes. The value of each pixel in the output image is based on a comparison of the corresponding pixel in the input image with its neighbors. By choosing the size and shape of the neighbourhood, you can construct a morphological operation that is sensitive to specific shapes in the input image. With GLCM (Gray Level Co-occurrence Matrix) the statistical parameters are obtained which provides the information about the texture of the image. Identification of the presence of worm. The response of image processing work using MATLAB is given in fig 2 to 13.

CONCLUSION:

The hookworm is a major intestinal parasites which causes serious problem to new born babies, women and persons who are malnourished. It is very difficult to identify the investigation of hookworm, in its initial stage. There are various technologies like PCR method, Mori Culture etc. It will be difficult to analyze the presence of hookworm. These technologies are less efficient when more number of images are analyzed at the same time, also initial identification of hookworm during its larval stage. When the number of images has been added to the database it is easily analyzed and segmented by using the proposed methodology. This method is even very effective when the collection of images has been collected by Wireless Capsule Endoscopic Method. The future work plan is to design a Wireless Capsule with low cost and user friendly implementation.

REFERENCES:

1. He, Jun-Yan, Et Al. "Hookworm Detection in Wireless Capsule Endoscopy Images with Deep Learning." IEEE Transactions on Image Processing. 27.5 (2018): 2379-2392.

2. Jaiswal, Priyanka K., and A. B. Kharate. "A Review: Efficient Classification of Wireless Capsule Endoscopy Images Using Artificial Neural Network." (2018).

3. Haritha, S., N. Nivetha, G. Radhika, and V. Jeyaramya. "Saliency Based Hookworm and Infection Detection for Wireless Capsule Endoscopy Diagnosis." (2017).

4. Sri, R. Sneha, A. Sheryl Oliver, and S. Shanthini. "Detection of Hookworm Infection in the Wireless Capsule Endoscopy Images." 2017 IEEE International Conference on Electrical, Instrumentation and Communication Engineering (ICEICE). IEEE, (2017).

5. Sasi, R., P. Sowmya, E. Vinodhini, and S. Shalini. "Hookworm Detection Using Wireless Capsule Endoscopy Images."(2017).

6. Tuba, Eva, Milan Tuba, and Raka Jovanovic. "An Algorithm for Automated Segmentation For Bleeding Detection In Endoscopic Images." International Joint Conference on Neural Networks (IJCNN). IEEE, (2017).

7. Yesodha, M., A. Angeline Nishidha, R. Ranjeetha, Cr Arthi Chandran, and S. Chitra. "Hookworm Detection in Wireless Capsule Endoscopy Images Using Edge Morphological Operator."(2017).

8. Saraswathy. N, Shalini. A, Annish Brislin. M.R, “Automatic Revelation Of Hookworm Based On Graph Cut Method From WCE Images”, (IJIRSET), May (2017).

9. Chen, Honghan, Junzhou Chen, Qiang Peng, Ganglu Sun, and Tao Gan. "Automatic Hookworm Image Detection for Wireless Capsule Endoscopy Using Hybrid Colour Gradient and Contourlet Transform." In 2013 6th International Conference on Biomedical Engineering and Informatics, Pp. 116-120. IEEE, (2013).

Received on 15.03.2019 Modified on 28.06.2019

Research J. Pharm. and Tech. 2020; 13(1):22-26.

DOI: 10.5958/0974-360X.2020.00004.9