CN-121999359-A - Tomato growth information monitoring method, device, equipment, medium and product

CN121999359ACN 121999359 ACN121999359 ACN 121999359ACN-121999359-A

Abstract

The application discloses a tomato growth information monitoring method, a tomato growth information monitoring device, tomato growth information monitoring equipment, a tomato growth information monitoring medium and a tomato growth information monitoring product, and relates to the technical field of intelligent farms. The method monitors the growth information based on the RGB image and the depth image, utilizes the RGB information in the RGB image and combines YOLOv target detection network model to realize the detection of targets (stem nodes and flowers), and utilizes the depth information of the depth image to realize the measurement of targets.

Inventors

XU XING
JIN WENQIN
ZHAO YUN
HE YONG
CHU BINGQUAN
WU NA

Assignees

浙江科技大学
浙江大学

Dates

Publication Date: 20260508
Application Date: 20241106

Claims (9)

1. A method for monitoring tomato growth information, comprising: acquiring RGB images and depth images of tomato planting areas; Aligning the depth image with the RGB image to obtain an aligned depth image; Inputting the RGB image into a YOLOv target detection network model after training, and detecting stem nodes and flowers of tomato plants in the RGB image to obtain a stem node detection frame and a flower detection frame; Intercepting a depth image corresponding to a detection frame of a stem node at the bottommost part of a tomato plant in the aligned depth image; according to the depth map corresponding to the detection frame of the stem node at the bottom, calculating the diameter of the stem of the tomato plant; According to the aligned depth images, determining the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower; and calculating the internode distance of the tomato plant according to the central depth pixel point of the detection frame of each stem node and the central depth pixel point of the detection frame of each flower of the same tomato plant.
2. The tomato growth information monitoring method according to claim 1, wherein aligning the depth image with the RGB image to obtain an aligned depth image, specifically comprises: projecting the depth image to a plane where the RGB image is positioned by using the following formula to obtain a projected depth image; Wherein, (u r ,v r ) is the pixel coordinates of the projected depth image, u r and v r are the abscissa and ordinate of the pixel of the projected depth image, respectively, (u d ,v d ,Z d ) is the pixel coordinates of the depth image, u d 、v d and Z d are the abscissa, ordinate and depth coordinate of the pixel of the depth image, respectively, K r is the reference matrix of the depth camera for obtaining the depth image, K d is the reference matrix of the RGB camera for obtaining the RGB image, and R and T are the rotation matrix and translation vector, respectively, for characterizing the relative pose between the depth camera and the RGB camera; Interpolation is carried out on the projected depth image by using the following formula, and an aligned depth image is obtained; Wherein Z (x, y) is an interpolation result at a pixel point (x, y) in the aligned depth image, x and y respectively represent an abscissa and an ordinate of the pixel point in the aligned depth image, Δx is a lateral distance of the pixel point (x, y) from a (i, j) th pixel point in the projected depth image, Δy is a longitudinal distance of the pixel point (x, y) from the (i, j) th pixel point in the projected depth image, i and j respectively represent a lateral position number and a longitudinal position number of a pixel in the projected depth image, Z (i, j) is a pixel value of the (i, j) th pixel point in the projected depth image, Z (i+1, j) is a pixel value of the (i+1, j) th pixel point in the projected depth image, and Z (i+1, j) is a pixel value of the (i+1).
3. The tomato growth information monitoring method according to claim 1, wherein the calculating of the diameter of the stem of the tomato plant according to the depth map corresponding to the detection frame of the bottommost stem node comprises: binarizing a depth map corresponding to a detection frame of the bottommost stem node to obtain a binarized depth map; Reconstructing the binarized depth map by using the following formula to obtain a binarized reconstructed image; Wherein, the Representing the pixel intensities of the pixel points (x ', y') in the binarized reconstructed image, x 'and y' representing the abscissa and ordinate, respectively, i x'y' representing the pixel intensities of the pixel points (x ', y') of the RGB image, Pixel intensities of pixel points (x ', y') representing overlapping portions of the RGB image and the binarized depth map, mean represents a mean value of the pixel intensities of the RGB image, σ thres represents a standard deviation of the pixel intensities of the RGB image; determining the number of pixels with the pixel value of 255 in the width range of a detection frame of the bottommost stem node in the binarized reconstructed image, and taking the number as the number of effective pixels; the product of the number of effective pixels and the width of a single pixel was calculated as the diameter of the stem of the tomato plant.
4. A tomato growth information monitoring method according to claim 3, wherein the binarization is performed on the depth map corresponding to the detection frame of the bottommost stem node, and the formula for obtaining the binarized depth map is: Wherein, the Representing depth values of pixel points (x ', y') in the binarized depth map, x 'and y' representing the abscissa and ordinate, respectively, D roi representing a depth threshold, A first fluctuation parameter is indicated and a second fluctuation parameter is indicated, Representing the depth value of the pixel point (x ', y') in the depth map corresponding to the detection frame of the bottommost stem node, Representing a second fluctuation parameter.
5. The tomato growth information monitoring method according to claim 1, wherein the calculating the pitch of the tomato plant according to the center depth pixel point of the detection frame of each stem node and the center depth pixel point of the detection frame of each flower of the same tomato plant specifically comprises: Judging whether a stem node of the tomato plant is shielded or not, and obtaining a judging result; If the judgment result is yes, calculating the distance between the center depth pixel point of the detection frame of the bottommost stem node of the same tomato plant and the center depth pixel point of the detection frame of the adjacent stem node as the pitch of the tomato plant according to the center depth pixel points of the detection frames of all stem nodes of the same tomato plant; If the judgment result is negative, calculating the distance between the central depth pixel point of the detection frame of the bottommost flower of the same tomato plant and the central depth pixel point of the detection frame of the adjacent flower according to the central depth pixel points of the detection frames of all the flowers of the same tomato plant, wherein the distance is used as the pitch distance of the tomato plant, and the adjacent flower is the flower which is positioned at the upper part of the bottommost flower and is adjacent to the bottommost flower.
6. A tomato growth information monitoring apparatus, characterized in that the tomato growth information monitoring apparatus applies the tomato growth information monitoring method according to any one of claims 1-5, the tomato growth information monitoring apparatus comprising: the image acquisition module is used for acquiring RGB images and depth images of the tomato planting area; an image alignment module, configured to align the depth image with the RGB image, and obtain an aligned depth image; The detection module is used for inputting the RGB image into a YOLOv target detection network model after training, detecting stem nodes and flowers of tomato plants in the RGB image, and obtaining a stem node detection frame and a flower detection frame; The depth map intercepting module is used for intercepting a depth map corresponding to a detection frame of a stem node at the bottommost part of the tomato plant in the aligned depth image; the stem diameter calculation module is used for calculating the diameter of the stems of the tomato plants according to the depth map corresponding to the detection frame of the stem node at the bottommost part; The central depth pixel point determining module is used for determining the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower according to the aligned depth images; And the internode distance calculation module is used for calculating the internode distance of the tomato plant according to the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower.
7. Computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the tomato growth information monitoring method according to any one of claims 1-5.
8. A computer readable storage medium having stored thereon a computer program, characterized in that the computer program, when executed by a processor, implements a method for monitoring tomato growth information as claimed in any one of claims 1-5.
9. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements a method for monitoring tomato growth information as claimed in any one of claims 1-5.

Description

Tomato growth information monitoring method, device, equipment, medium and product Technical Field The application relates to the technical field of intelligent farms, in particular to a tomato growth information monitoring method, a tomato growth information monitoring device, tomato growth information monitoring equipment, tomato growth information monitoring medium and tomato growth information monitoring products. Background The plant growth information is reliable data reflecting the current growth stage of the plant and is important for the establishment of plant management measures. Therefore, accurate measurement of plant growth information is of great importance for the management and high yield of intelligent farms. The method for acquiring the growth information of the greenhouse tomato plants is usually a manual measurement method, however, the manual measurement method has the problems of high labor intensity, high cost, inconsistent results, large measurement error and the like. Disclosure of Invention The application aims to provide a tomato growth information monitoring method, a device, equipment, a medium and a product, so as to realize automatic monitoring of tomato growth information, save labor and cost and improve the monitoring efficiency and consistency and accuracy of monitoring results. In order to achieve the above object, the present application provides the following. In a first aspect, the present application provides a method for monitoring tomato growth information, which is characterized by comprising: And acquiring RGB images and depth images of the tomato planting area. And aligning the depth image with the RGB image to obtain an aligned depth image. And inputting the RGB image into a YOLOv target detection network model after training, and detecting stem nodes and flowers of the tomato plants in the RGB image to obtain a stem node detection frame and a flower detection frame. And intercepting a depth map corresponding to a detection frame of the bottommost stem node of the tomato plant in the aligned depth image. And calculating the diameter of the stems of the tomato plants according to the depth map corresponding to the detection frame of the bottommost stem node. And determining the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower according to the aligned depth images. And calculating the internode distance of the tomato plant according to the central depth pixel point of the detection frame of each stem node and the central depth pixel point of the detection frame of each flower of the same tomato plant. In a second aspect, the present application provides a tomato growth information monitoring apparatus, the tomato growth information monitoring apparatus applying the above-mentioned tomato growth information monitoring method, the tomato growth information monitoring apparatus comprising: and the image acquisition module is used for acquiring RGB images and depth images of the tomato planting area. And the image alignment module is used for aligning the depth image with the RGB image to obtain an aligned depth image. And the detection module is used for inputting the RGB image into the trained YOLOv target detection network model, and detecting stem nodes and flowers of the tomato plants in the RGB image to obtain a stem node detection frame and a flower detection frame. And the depth map intercepting module is used for intercepting the depth map corresponding to the detection frame of the bottommost stem node of the tomato plant in the aligned depth image. And the stem diameter calculation module is used for calculating the diameter of the stems of the tomato plants according to the depth map corresponding to the detection frame of the bottommost stem node. And the central depth pixel point determining module is used for determining the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower according to the aligned depth images. And the internode distance calculation module is used for calculating the internode distance of the tomato plant according to the central depth pixel point of the detection frame of each stem node of the same tomato plant and the central depth pixel point of the detection frame of each flower. In a third aspect, the application provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the computer program to implement the tomato growth information monitoring method described above. In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the tomato growth information monitoring method described above. In