Method for identifying ripeness, computing device, and ripeness identification system

Abstract

A maturity identification method includes: capturing images of a fruit to be tested using a camera device to obtain a plurality of image data; and performing the following steps using a computer device: performing a bottom detection operation on the fruit to be tested based on the image data and a detection model to mark a first attention area of ​​one of the image data and a second attention area of ​​the other image data; and performing a maturity identification operation on the fruit to be tested based on the first attention area or the second attention area and the detection model, wherein the maturity identification operation includes: performing a color identification operation; performing a first wilting rate detection operation when the color of the bottom of the fruit to be tested is a preset color; and outputting a maturity result when the first detection result of the first wilting rate detection operation is negative.

Claims

1. A maturity identification method, comprising: A plurality of image data are acquired by capturing images of a fruit subject at a specific frequency using a camera device; and the following steps are performed by a computer device: Based on the image data and a detection model, a bottom detection operation of the fruit subject is performed to mark a first attention area of ​​one of the image data and a second attention area of ​​the other of the image data; and based on the first attention area or the second attention area and the detection model, a maturity identification operation of the fruit subject is performed, and a maturity result is output, wherein the maturity identification operation includes: performing a color identification operation, the color identification operation including detecting whether the color of the bottom of the fruit subject in the first attention area is a preset color; When the color of the bottom of the fruit to be tested is the preset color, a first withering rate detection operation is performed. The first withering rate detection operation includes detecting whether a first withering rate of a plurality of flower bracts of the fruit to be tested is within a first preset range, so as to generate a first detection result; and when the first detection result of the first withering rate detection operation is negative, the generated maturity result is a first maturity result or a second maturity result, wherein the first maturity result includes a first maturity grade, the first attention area and the image data, and the second maturity result includes a second maturity grade, the first attention area and the image data.

2. The maturity identification method as described in claim 1 further includes performing the following steps via the computer device: performing a pre-training operation on the detection model based on a plurality of training image data, the pre-training operation including: Perform a data augmentation operation, which includes increasing the amount of training image data through image processing; perform an image annotation operation, which includes annotating the bottom of the fruit test object and the flower bracts in the training image data using an image annotation tool; perform an image conversion operation, which includes converting the coordinates of the bottom and the flower bracts into a JSON file format; and perform a storage operation, which includes inputting the JSON file format and the augmented image data into the detection model.

3. The maturity identification method as described in claim 2, wherein the bottom detection operation includes: Based on each of these image data, the bottom of the fruit to be tested is detected; A bounding box is established at the bottom of the fruit to be tested; and the center point of the bounding box is detected to be within a second preset range to generate the first attention area and the second attention area.

4. The maturity identification method as described in claim 3, wherein when the color of the bottom of the fruit to be tested is not the preset color, the resulting maturity result is a third maturity result, wherein the third maturity result includes a third maturity grade, the first attention area, and the image data.

5. The maturity identification method as described in claim 3, wherein when the first detection result of the first wilting rate detection operation is yes, the computer device performs a second wilting rate detection operation based on the other of the image data.

6. The maturity identification method as described in claim 5, wherein the second wilting rate detection operation includes: Calculate a second withering rate of the flower bracts of the fruit to be tested in the second attention area; calculate an average withering rate of the first withering rate and the second withering rate; and detect whether the average withering rate is greater than a preset value to generate a second detection result.

7. The maturity identification method as described in claim 6, wherein when the second detection result of the second withering rate detection operation is yes, the generated maturity result is a fourth maturity result, wherein the fourth maturity result includes the first maturity grade, the second attention area, and the other of the image data.

8. The maturity identification method as described in claim 7, wherein when the second detection result of the second withering rate detection operation is negative, the generated maturity result is a fifth maturity result, wherein the fifth maturity result includes the second maturity grade, the second attention area, and the other of the image data.

9. A computer device for identifying the ripeness of a fruit sample, the computer device comprising: A memory module for storing multiple image data of the fruit to be tested; The system also includes a processor electrically connected to the memory, which receives the image data and performs the following steps: performing a bottom detection operation on the fruit to be tested based on the image data and a detection model to mark a first attention area of ​​one of the image data and a second attention area of ​​the other; and performing a maturity identification operation on the fruit to be tested based on the first attention area or the second attention area and the detection model, and outputting a maturity result, wherein the maturity identification operation includes: performing a color identification operation, the color identification operation including detecting whether the color of the bottom of the fruit to be tested in the first attention area is a preset color; When the color of the bottom of the fruit to be tested is the preset color, a first withering rate detection operation is performed. The first withering rate detection operation includes detecting whether a first withering rate of a plurality of flower bracts of the fruit to be tested is within a preset range, so as to generate a first detection result; and when the first detection result of the first withering rate detection operation is negative, the generated maturity result is a first maturity result or a second maturity result, wherein the first maturity result includes a first maturity grade, the first attention area and the image data, and the second maturity result includes a second maturity grade, the first attention area and the image data.

10. A maturity identification system, comprising: A camera device is used to capture images of a fruit to be tested at a specific frequency in order to obtain multiple image data. The system also includes a computer device electrically connected to the camera device, wherein the computer device is configured to perform the following steps: performing a bottom detection operation on a fruit sample based on the image data and a detection model, to mark a first attention area of ​​one of the image data and a second attention area of ​​the other of the image data; and performing a maturity identification operation on the fruit sample based on the first attention area or the second attention area and the detection model, and outputting a maturity result, wherein the maturity identification operation includes: performing a color identification operation, the color identification operation including detecting whether the color of the bottom of the fruit sample in the first attention area is a preset color; When the color of the bottom of the fruit to be tested is the preset color, a first withering rate detection operation is performed. The first withering rate detection operation includes detecting whether a first withering rate of a plurality of flower bracts of the fruit to be tested is within a first preset range, so as to generate a first detection result; and when the first detection result of the first withering rate detection operation is negative, the generated maturity result is a first maturity result or a second maturity result, wherein the first maturity result includes a first maturity grade, the first attention area and the image data, and the second maturity result includes a second maturity grade, the first attention area and the image data.

Images