Penganalisis Kesegaran Daging Sapi Dan Daging Babi Mentah Berdasarkan Klasifikasi Warna Dan Kelembaban

Abstract

There are many frauds involving the trade of beef. Traders do the fraud by replacing the beef it sells with pork. Also, the lack of consumer knowledge of the difference between beef and pork make these traders unscrupulous. Because we research by designing and creating systems that can identify and analyze the freshness of beef and pork. This system uses a color sensor as a color identifier in meat and humidity sensors to measure the level of freshness of meat. Also used is also a microcontroller as an acquirer and data processor. This system is designed to distinguish between beef and pork. This study involved data collection on beef and pork as much as 420 data. These data have shown that this system has an accuracy of 79%, 80%, 80%, 83%, 77% and 81% in identifying fresh beef, fresh pork, less fresh beef, less fresh pork, the cow is not fresh, and the pork is not fresh in sequence. This system is portable because it can be taken anywhere with a battery that supplies the power of this tool. With this system, consumers can distinguish between beef and pork and can know the level of freshness in each meat.

Fuzzy Learning Vector Quantization for Classification of Mixed Meat Image Based on Character of Color and Texture

“…𝑥 ̃= (𝑥 (1) , 𝑥 (2) , 𝑥 (3) ) (33) where 𝑥 (1) is the minimum value, 𝑥 (2) is the value of the midpoint of the peak, and 𝑥 (3) is the maximum value [24]. To deal with the fuzziness caused by errors in statistical measurements, FLVQ uses fuzzy numbers to determine the activity of these neurons (Figure 9).…”

“…The FLVQ algorithm has 3 conditions in the weight update [24]. First, there is no winning codebook because all the maximum similarity values are 0 so the data is considered not to fall into any class (equations ‫)1(𝑗𝑖𝑊‬ = 𝑊 𝑖𝑗 (2) − 𝛽(𝑊 𝑖𝑗 (2) − 𝑊 𝑖𝑗 (1) )…”

“…Second, the codebook with the maximum similarity value belongs to the same class as the input vector so that the codebook is closer to the input vector and produces higher similarity. Calculations are made based on equation ‫)2(𝑗𝑖𝑊(‬ = 𝑊 𝑖𝑗 (2) + 𝛼(1 − 𝜇 𝑖𝑗 )(× 𝑖 − 𝑊 𝑖𝑗 (2) ) (38).…”

“…Meat is an important food element in addressing nutritional demands since it contains high-quality protein and full amino acids [1]. Meat comes in many varieties, including beef, pork, mutton, poultry, and so on [2]. Beef is consumed more frequently because it is high in nutrients and animal protein, as well as is easy to cook [3].…”

“…Pork is paler than beef in color. Meanwhile, beef fiber is more evident than pig fiber in terms of texture [2]. One way that can be used to recognize beef and pork in the field of informatics is to use digital image techniques considering that digital image processing can create a computerized meat recognition system that can distinguish beef and pork like humans [8].…”