Unique Registration Number: GSIC- 308
Innovator Name:
B. Pavan Kalyan
Siddhant Kunde
Vaishabh Jalmi
Contact No: +918766473081
Project Objective:
- To classify different fish species using image processing and machine learning
techniques. - To predict the freshness of fish in terms number of days after catch using image
processing and machine learning techniques. - To design a non-Invasive hardware/software system to detect presence of formalin in
fish and its concentration. - To design and develop an end-to-end system on android which performs above
three objectives
Contact Email: kalyanpavan2299@gmail.com
Abstract:
Fish is most popularly consumed in coastal areas of India and some hinterlands as well. The demand for fish has been increasing every year. To cater this demand fish has to be transported to many metro cities and tourism dominated areas. Which leads to illegal increase of adulteration of fish using chemicals (formaldehyde) for preservation. This in turn leads to several health issues, and also it is one of the major component that causes cancer. This project presents an approach to analyze the quality of fish. The overall objective is to create a complete system on a mobile phone, with the combination of various software techniques as well as a hardware interface
Goa State Innovation Council :
There are many methods to detect formalin both invasive and non-invasive. One such method is testing through pH strips from formalin detection kit. This kit is difficult for publics use. Fish quality analysis system, which is the project outcome will be able to detect artificial formalin on fish. This system is easy to use. The system, consists of HCHO sensor to detect the presence of artificial formalin on fish and uses image processing and machine learning techniques to classify fish into different family types and predict the fish freshness based on number of days after the catch.
Project Outcome/result/findings:
Objective 1 outcome:
Setup for image dataset generation
Innovative Approach: (in 50 words)
Most of methods used for detection of formalin on fish are invasive system. This project aims developed a non-invasive system. Certain testing methods make use of chemical reagent for formalin detection which creates a problem for common people to handle chemicals. This project avoids this drawback as no chemicals are involved. Many times a non-local person is not able to identify a local fish and also not gauge the freshness of fish in terms of days from the catch. This system identifies the type/class/family of fish and also freshness of fish. Finally, there is a requirement for a complete system which will perform all above analysis and give a one stop solution to the fish consumers and retailers.