Published Paper Details:

In this work we demonstrate the design methodology for An Intelligent Fish freshness system for real time classification of the freshness of fish using Artificial Neural Network (ANN). Automated Fish freshness assessment and identification plays an important role in fisheries industry applications. This method is based on the series of sensor connected together with Arduino At mega-328 to improve the performance of fish freshness quality assessment. The result is used for the fish freshness using artificial neural network (ANN). The series of sensor node design involves three major parts (1) the gas sensing elements i.e. (sensors) (ii) A Single board microcontroller unit (Arduino) (iii) Artificial Neural Network. The Sensor interfacing methodology has been demonstrated in practice, by considering MQ series (MQ-4,MQ-2,MQ-8,MQ-7,MQ-5 and MQ-135) of Gas sensors elements, With the single board microcontroller i.e. Arduino Uno, in our case. Once the series of gas sensors has been interfaced with the single board microcontroller, the signals from the physical gas sensors are captured by running a software PLX-DAQ. Data was collected from three/3 selected species (i)Tilapia Fish (ii) Carpio Fish and (iii) Tengra Fish over a period of several days. The ANN was trained with many samples as collected while testing was done only by using fresh fish data (Day 1), Semi-spoiled Fish (Day 2) and spoiled fish data (Day 3) by using 9 samples of three different species. The entire test samples were classified with 99% accuracy. The classified output has a mean square error of 1.4155�10^-17 at 9022 epochs for Tilapia, 5.9805�10^-13 at 9448 epochs for Carpio and 5.925�10^-18 at 10,000 epochs for Tengra. Intelligent processing of the sensor patterns involves the use of a dedicated ANN for each species under study. The ANN has been trained in NCC laboratory conditions with ensure reliability, accuracy and repeatability. The proposed system has been successful in identifying the number of days after catching the fish with an accuracy of up to 99%.