Prediction device, prediction method, program, and prediction system

Abstract

The present invention generates an AI for predicting progression.　Provided is a training device comprising: an imaging unit that acquires an image of cooking oil; a state identification unit that analyzes the image to identify a state of the cooking oil; and a learning unit that causes a learning model to learn a correlation between the state and a progression indicating the degree to which the cooking oil has degraded.

Claims

1. An imaging unit that acquires images of cooking oil, A state identification unit analyzes the aforementioned image to identify the state of the edible oil, A learning unit that trains a learning model on the correlation between the aforementioned state and the degree of deterioration of the edible oil, The system includes a prediction unit that predicts the progress using a trained model trained on the aforementioned learning model, The aforementioned state is the rate at which bubbles are generated in the edible oil per predetermined time, the size of the bubbles, the rate at which the bubbles disappear, or a combination thereof. Or, The aforementioned progression is, The viscosity increase rate of the edible oil, the color of the edible oil, the amount of polar compounds in the edible oil, the smoke point of the edible oil, the tocopherol content of the edible oil, the iodine value of the edible oil, the amount of volatile components in the edible oil, the composition of volatile components in the edible oil, the flavor of the edible oil, the amount of volatile components in the fried food fried in the edible oil, the composition of volatile components in the fried food, the flavor of the fried food, or a combination thereof. Based on the progress, the system further outputs a prediction of the types or number of fried foods that can be cooked by the time the cooking oil needs to be changed. Prediction device.

2. A prediction device according to claim 1, The aforementioned state is, A prediction device that further includes the number of bubbles.

3. A prediction device according to claim 1 or 2, The aforementioned progression is, A prediction device further comprising the acid value of the edible oil, the anisidine value of the edible oil, the carbonyl value of the edible oil, or the refractive index of the edible oil.

4. A prediction device according to any one of claims 1 to 3, The aforementioned state is, A prediction device that identifies the initial foam generated in edible oils with an acid value of less than 0.

2.

5. A prediction device according to any one of claims 1 to 4, The aforementioned state is, The size of the bubbles that form in the aforementioned edible oil is further included, The bubbles are classified into first bubbles based on their size, and second bubbles that are smaller than the first bubbles. Based on the first bubble and the second bubble, the progress is predicted. Prediction device.

6. A prediction device according to claim 5, The size of the bubbles that form in the edible oil is determined by extracting edge components from the image, extracting circular or elliptical shapes in the image using image recognition, and then determining the diameter or circumference of the bubbles. Prediction device.

7. A prediction method performed by a prediction device, The prediction device performs an imaging procedure to acquire an image of the edible oil, A prediction device analyzes the image to determine the state of the edible oil, and A learning procedure for a prediction device to train a learning model on the correlation between the state and the degree of deterioration of the edible oil, The prediction device has a prediction procedure that predicts the progress using the trained model that has been trained on the learning model, The aforementioned state is the rate at which bubbles are generated in the edible oil per predetermined time, the size of the bubbles, the rate at which the bubbles disappear, or a combination thereof. Or, The aforementioned progression is, The viscosity increase rate of the edible oil, the color of the edible oil, the amount of polar compounds in the edible oil, the smoke point of the edible oil, the tocopherol content of the edible oil, the iodine value of the edible oil, the amount of volatile components in the edible oil, the composition of volatile components in the edible oil, the flavor of the edible oil, the amount of volatile components in fried food fried in the edible oil, the composition of volatile components in the fried food, the flavor of the fried food, or a combination thereof. Prediction method.

8. A program for causing a computer to execute the prediction method described in claim 7.

9. A prediction system having a prediction device according to any one of claims 1 to 6.

Images