Fish Quality Assessment using smart phones

Currently, the rising interest of people in food quality has contributed to the growing demand for quick systematic technologies. Different electrical and sensory techniques have been used to assess the quality of fish. However, these techniques are timing consuming, costly and may be invasive and d

Project Title

Fish Quality Assessment using smart phones

Project Area of Specialization

Artificial Intelligence

Project Summary

Currently, the rising interest of people in food quality has contributed to the growing demand for quick systematic technologies. Different electrical and sensory techniques have been used to assess the quality of fish. However, these techniques are timing consuming, costly and may be invasive and destructive. In today era, image processing is widely used in biodiversity applications. After limited time, consuming the fish can harm people health. Fish also start involving safety aspects such as being free from harmful bacteria, parasites or chemicals. In this paper, a very simple new-fangled way of deterring the fish freshness has been proposed by considering three diverse features of the fish. Two of the features are related to shape and one is to colour. The proposed approach has been tested on a dataset consisting of 100 images that are captured at different time in a week. Experimental results depict that using the shape, 93 percent of the tested fishes have been precisely recognised while using colour features, all the tested fishes have been successfully classified. These results prove that the proposed technique can be easily and successfully used to detect the freshness of the fish. The system was developed in Matlab R2018b with Intel core i5, 2.24 GHz processor having 8 GB RAM.

Project Objectives

To design and formulate an automatic system which automatically detect fish quality using mobile cam.

Project Implementation Method

In this research work, a very simple new-fangled way of deterring the fish freshness has been proposed by considering three diverse features of the fish. Two of the features are related to shape and one is to color. The proposed fish quality assessment system consists of four phases which has been discussed below.

The first phase of our proposed system is the image acquisition. We acquired all the fish (Sultan) images using Samsun mobile S5 and the collected images are of good quality. We don’t require any additional lighting equipment. We took two (2) photographs of each and every fish to analyze well our proposed system. The classification is based on fresh and stale fish. Therefore, photographs were taken at twenty-four (24) hour time interval where the first photo shows the fresh while the last represents the stale fish. The second phase of our proposed system is pre-processing. Normally, camera image contains much more information and it is difficult and time consuming to processed. Therefore, we first used some pre-processing techniques to obtain only useful information. So, we used grayscale conversion, binary gradient mask, dilation and removing redundant image parts in this phase. Next section shows all the images details. The third step belongs to feature extraction. In order to get identify the quality or freshness of the fish, we much need some basic feature to extract from the pre-processed images. First of all, the black and white image is analyzed and fines the top and bottom white pixels to get the contours of the fish. Using this way, we can also minimize the constant position requirement of the fish. To assess the freshness, we calculate three (3) types of features. Firstly, we calculate the five different regions of interest and find the average value of all those regions. Using this way, we assessed the fish freshness in a simple and robust way. Secondly, we calculate the slope of five different regions and we observed that individual region is not enough to provide clear separation. Therefore, we group the regions to get clear separation between fresh and the stale fish. The third way is to get color (Hue) information. Using Hue information, we can easily identify freshness of the fish. The details with figures and photographs are presented in next section.

Benefits of the Project

This system will use a very simple new-fangled way of deterring the fish freshness has by considering three diverse features of the fish. Two of the features are related to shape and one is to color. The proposed system can have a great impact on society as well as it will improve our food quality easily.

Technical Details of Final Deliverable

this system will be an an android app which will automaticaly  assess the quality of a fish in an efficient way. A layman will use this app in a simple way to check the fish details. 

Final Deliverable of the Project

Hardware System

Core Industry

IT

Other Industries

Health

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com