A method for online detection of rice lodging direction information

Abstract

The present invention provides an online detection method for rice lodging direction information, comprising the following steps: pre-processing an acquired lodging rice image, segmenting the lodging rice image into a feature region to be identified according to grayscale, and extracting the contour of the lodging region after morphologically enhancing the feature region to be identified; creating a grayscale co-occurrence matrix of the contour region according to the contour of the lodging region, calculating the grayscale value characteristics of the contour region, and obtaining rice texture information; extracting local edge direction lines through regionalization and block division after the rice texture information is texture enhanced, thereby determining the rice lodging direction. The present invention can quickly and accurately extract the lodging direction information of small-scale lodging rice, and the extraction effect is good.

Description

Technical Field

[0001] The present invention relates to the field of intelligent agricultural machinery and equipment or the field of image recognition technology, and in particular to an online detection method for rice lodging direction information. Background Art

[0002] As one of the three major food crops in the world, rice is second only to wheat in terms of cultivated area and total output. my country's total rice output ranks first in the world, and stable rice production is inseparable from national food security. Rice lodging refers to the phenomenon that large tracts of upright rice tilt or even lie flat on the ground due to natural factors or external forces. Lodging will have a great impact on rice yield and quality, and will also lead to difficulties in harvesting. Real-time and accurate acquisition of small-scale proximal lodging crop characterization information is of great significance for the intelligent real-time harvesting of lodging crops by combine harvesters.

[0003] Existing lodging monitoring research is divided into monitoring based on satellite spectroscopy, radar, drones, and multi-source data fusion. Satellite remote sensing covers a large area, and the spectral changes caused by lodging are weak, making it difficult to clearly present lodging information; due to the uncertainty of the lodging area and interference from the surrounding environment, radar cannot accurately monitor; drone platforms equipped with digital cameras or spectral imagers are susceptible to multiple factors such as changes in altitude angles during image acquisition, causing lodging information to be lost. At present, crop lodging measurements are mostly large-scale measurements that only detect whether lodging has occurred, and are mostly used for lodging disaster assessment and yield improvement. However, there are few studies that use proximal sensing to monitor small-scale lodging in real time and efficiently, and extract detailed lodging crop characterization information for agricultural practice. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an online detection method for rice lodging direction information, which can quickly and accurately extract the lodging direction information of rice that has fallen in a small area, and the extraction effect is good.

[0005] The present invention achieves the above technical objectives through the following technical means.

[0006] A method for online detection of rice lodging direction information comprises the following steps:

[0007] Preprocessing the acquired lodging rice image, segmenting the lodging rice image into feature areas to be identified according to grayscale, and extracting the contour of the lodging area after morphologically enhancing the feature areas to be identified;

[0008] A grayscale co-occurrence matrix of the contour area is created according to the contour of the lodging area, and the grayscale value characteristics of the contour area are calculated to obtain rice texture information; after the rice texture information is texture enhanced, local edge direction lines are extracted through regionalized blocks to determine the rice lodging direction.

[0009] Furthermore, the lodging rice image obtained by the preprocessing is specifically:

[0010] The region of interest is extracted from the acquired lodging rice image, and the channel of the region of interest is separated into three single-channel images of R, G, and B; the image with the most prominent contour features in the single-channel image is selected for mean filtering.

[0011] Furthermore, the lodging rice image is segmented into feature regions to be identified according to grayscale, and the contour of the lodging region is extracted after the feature regions to be identified are enhanced by morphology, specifically:

[0012] The preprocessed region of interest is segmented using a grayscale threshold, and the segmented region is eroded with a rectangular structure element; the eroded region is then connected and filled; the filled image is closed with a rectangular structure element, and the image after the closing operation is refilled; the refilled image is eroded with a circular structure element; the largest area after the erosion is selected, and the largest area after the erosion is closed with a rectangular structure element, the area feature threshold is set, and the characteristic area is segmented; the characteristic area is closed with a rectangular structure element, and the sub-pixel edge contour is extracted from the characteristic area after the closing operation to obtain the contour of the lodging area.

[0013] Furthermore, the grayscale co-occurrence matrix of the contour area is created according to the contour of the lodging area, and the grayscale value characteristics of the contour area are calculated to obtain the rice texture information, specifically:

[0014] The maximum inscribed rectangle is determined according to the outline of the lodging area, and the image within the maximum inscribed rectangular area is transformed into grayscale; a grayscale co-occurrence matrix is ​​created in the image area after the grayscale transformation, and the adjacent frequency values ​​are determined according to the grayscale pixel values ​​in the maximum inscribed rectangular area after the grayscale transformation and added to the co-occurrence matrix; the grayscale eigenvalue of the texture is calculated through the co-occurrence matrix, and a texture filter is formed with the grayscale eigenvalue of the texture to filter the image area after the grayscale transformation to obtain the rice texture information.

[0015] Furthermore, the rice texture information is processed by texture enhancement and local edge direction lines are extracted through regionalization and block division to determine the rice lodging direction, specifically:

[0016] The rice texture information is enhanced, the texture information in the enhanced image area is evenly divided, the linear equation of the sub-pixel edge lodging flow line of each image area after the even division is extracted, and the rice lodging direction is statistically determined.

[0017] Furthermore, the texture enhancement process is: enhancing the rice texture information through Gabor filtering to highlight the linear texture.

[0018] Furthermore, the sub-pixel edges of each image area are binarized to obtain a two-dimensional point set; the least squares method is used to fit the straight line to minimize the vertical error from each point to the straight line, and the straight line equation of the lodging flow line is calculated.

[0019] Furthermore, the rice lodging direction is determined according to the slope of the linear equation of the lodging flow line.

[0020] Furthermore, a binocular camera is installed above the combine harvester cab, and the binocular camera lens is inclined at an acute angle to the horizontal line to obtain images of lodged rice.

[0021] A system for an online detection method for rice lodging direction information comprises a recognition system, wherein the recognition system stores a program of the online detection method for rice lodging direction information.

[0022] The beneficial effects of the present invention are:

[0023] Compared with the existing rice lodging monitoring technology, most of which is limited to predicting whether rice lodging will occur, and most of them are delayed predictions, the present invention can not only monitor the occurrence of rice lodging in real time, but also extract the contour of rice lodging in real time, and further extract the characterization information of rice lodging based on the contour, which can well adapt to the real-time recognition requirements of the characterization information of lodging rice during the intelligent mechanized harvesting of lodging rice. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. The drawings described below are some embodiments of the present invention. For ordinary technicians in this field, it is obvious that other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 This is a flow chart of the online detection method for rice lodging direction information of the present invention.

[0026] Figure 2 Schematic diagram of the installation of the binocular camera described in the present invention

[0027] Figure 3 This is the original picture of fallen rice.

[0028] Figure 4 This is the pre-processing map of the lodging area.

[0029] Figure 5 This is a morphological processing diagram of the lodging contour.

[0030] Figure 6 Extract the lodging texture map.

[0031] Figure 7Filter map for texture filter.

[0032] Figure 8 This is the extraction diagram of the straight line equation of the lodging direction. DETAILED DESCRIPTION

[0033] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0034] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0035] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "axial", "radial", "vertical", "horizontal", "inner", "outer" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0036] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0037] like Figure 1 As shown, the online detection method for rice lodging direction information of the present invention comprises the following steps:

[0038] S01: Figure 2As shown in the figure, a binocular camera is mounted above the cab of a combine harvester, with the lens tilted downward at a 30° angle to the horizontal line to collect left and right three-channel images of lodging rice with a resolution of 698×392, which is used to obtain images of lodging rice.

[0039] S02: Extract the region of interest from the acquired lodging rice image, separate the region of interest channel into three single-channel images of R, G, and B; select the image with the most prominent contour features in the single-channel image for mean filtering, such as Figure 3 As shown, the contrast can be increased, edge details can be highlighted, and noise can be reduced again.

[0040] S03: Use grayscale threshold to segment the preprocessed region of interest, and use rectangular structural elements to erode the segmented region; for regions of interest with uneven grayscale value distribution, it is necessary to expand the upper and lower limits of the threshold segmentation, which reduces the adaptability of the threshold segmentation to different lodging images. It is necessary to add lighting preprocessing operations instead of adjusting the threshold segmentation, such as Figure 4 shown.

[0041] Then, the eroded area is connected and filled; the filled image is closed with a rectangular structure element, and the closed image is refilled; the refilled image is eroded with a circular structure element; the largest area after erosion is selected, and the largest area after erosion is closed with a rectangular structure element, and the area feature threshold is set to segment and obtain the feature area; the feature area is closed with a rectangular structure element, such as Figure 5 As shown in the figure, the sub-pixel edge contour is extracted from the characteristic area after the closing operation to obtain the contour of the lodging area, as shown in Figure 6 shown.

[0042] S04: Determine the maximum inscribed rectangle according to the outline of the lodging area, and perform grayscale transformation on the image within the maximum inscribed rectangular area; create a grayscale co-occurrence matrix in the image area after the grayscale transformation, determine the adjacent frequency values ​​according to the grayscale pixel values ​​in the maximum inscribed rectangular area after the grayscale transformation and add them to the co-occurrence matrix; calculate the grayscale eigenvalues ​​of the texture through the co-occurrence matrix, and the grayscale eigenvalues ​​of the texture include texture energy, correlation, local uniformity and contrast; form a texture filter with the grayscale eigenvalues ​​of the texture to filter the image area after the grayscale transformation to obtain rice texture information.

[0043] S05: Enhance the rice texture information by using Gabor filtering to highlight the linear texture. Figure 7 shown.

[0044] S06: Equalize the texture information in the enhanced image area, extract the linear equation of the lodging flow line of the sub-pixel edge of each image area after equalization, and statistically determine the rice lodging direction. Binarize the sub-pixel edge of each image area to obtain a two-dimensional point set; use the least squares method to fit the straight line, minimize the vertical error from each point to the straight line, and calculate the linear equation of the lodging flow line. The rice lodging direction is determined according to the slope of the linear equation of the lodging flow line, such as Figure 8 shown.

[0045] A system for an online detection method for rice lodging direction information, characterized in that it comprises a recognition system, wherein the recognition system stores a program of the online detection method for rice lodging direction information.

[0046] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0047] The series of detailed descriptions listed above are only specific descriptions of feasible embodiments of the present invention. They are not intended to limit the scope of protection of the present invention. All equivalent embodiments or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for online detection of rice lodging direction information, characterized in that: The steps include: The acquired lodging rice image is preprocessed, the lodging rice image is segmented into the feature area to be identified according to the gray level, and the contour of the lodging area is extracted after the feature area to be identified is enhanced by morphology, specifically: The preprocessed region of interest is segmented using the grayscale threshold, and the segmented region is eroded using the rectangular structure element; the eroded region is then subjected to a connected domain operation and a filling operation; the filled image is closed using the rectangular structure element, and the closed image is refilled; The refilled image is corroded with a circular structure element; the largest area after corrosion is selected, and the largest area after corrosion is closed with a rectangular structure element, and the area feature threshold is set to segment and obtain the characteristic area area; the characteristic area area is closed with a rectangular structure element, and the sub-pixel edge contour is extracted from the characteristic area area after the closing operation to obtain the contour of the lodging area; A grayscale co-occurrence matrix of the contour area is created according to the contour of the lodging area, and the grayscale value characteristics of the contour area are calculated to obtain rice texture information; after the rice texture information is texture enhanced, local edge direction lines are extracted through regionalized blocks to determine the rice lodging direction.

2. The method for online detection of rice lodging direction information according to claim 1, characterized in that: The lodging rice image obtained by the preprocessing is specifically: The region of interest is extracted from the acquired lodging rice image, and the channel of the region of interest is separated into three single-channel images of R, G, and B; the image with the most prominent contour features in the single-channel image is selected for mean filtering.

3. The method for online detection of rice lodging direction information according to claim 1, characterized in that: The grayscale co-occurrence matrix of the contour area is created according to the contour of the lodging area, and the grayscale value characteristics of the contour area are calculated to obtain the rice texture information, specifically: Determine the maximum inscribed rectangle according to the contour of the lodging area, and perform grayscale transformation on the image within the maximum inscribed rectangle area; create a grayscale co-occurrence matrix in the image area after the grayscale transformation, and determine the adjacent frequency values ​​according to the grayscale pixel values ​​in the maximum inscribed rectangle area after the grayscale transformation and add them to the co-occurrence matrix; The grayscale eigenvalue of the texture is calculated through the co-occurrence matrix, and the grayscale eigenvalue of the texture is used to form a texture filter to filter the image area after grayscale transformation to obtain the rice texture information.

4. The method for online detection of rice lodging direction information according to claim 3, characterized in that: After the rice texture information is processed by texture enhancement, local edge direction lines are extracted through regionalization and block division, so as to determine the rice lodging direction, specifically: The rice texture information is enhanced, the texture information in the enhanced image area is evenly divided, the linear equation of the sub-pixel edge lodging flow line of each image area after the even division is extracted, and the rice lodging direction is statistically determined.

5. The method for online detection of rice lodging direction information according to claim 4, characterized in that: The texture enhancement process is: enhancing the rice texture information through Gabor filtering to highlight the linear texture.

6. The method for online detection of rice lodging direction information according to claim 4, characterized in that: The sub-pixel edges of each image area are binarized to obtain a two-dimensional point set. The least squares method is used to fit the straight line to minimize the vertical error from each point to the straight line, and the straight line equation of the lodging flow line is calculated.

7. The method for online detection of rice lodging direction information according to claim 4, characterized in that: The rice lodging direction is determined according to the slope of the linear equation of the lodging flow line.

8. The method for online detection of rice lodging direction information according to claim 1, characterized in that: A binocular camera is installed above the combine harvester cab, and the binocular camera lens is inclined at an acute angle to the horizontal line, so as to obtain images of lodged rice.

Images