技术领域Technical field
本发明涉及食用菌菌丝检测技术领域，具体地说，涉及一种食用菌菌丝长势检测方法及装置。The present invention relates to the technical field of edible fungus hyphae detection, and specifically to a method and device for detecting the growth of edible fungus hyphae.
背景技术Background technique
在食用菌的育种、种植等过程中，需要对食用菌的菌丝长势进行测定，以筛选出优良菌种或对食用菌的生长状态进行监测。对菌丝长势的测定，主要是对菌丝的生长速度、密实度、壮实度等参数的测定。常规测定手段，大多是依靠人工测量菌落尺寸、观察菌落密集度和色泽形态等方式实现。此种手段因观测者的主观因素难以客观科学，且效率较为低下。In the process of breeding and planting edible fungi, it is necessary to measure the mycelium growth of edible fungi in order to screen out excellent strains or monitor the growth status of edible fungi. The measurement of mycelium growth mainly involves the measurement of parameters such as growth rate, density, and robustness of mycelium. Conventional measurement methods mostly rely on manual measurement of colony size, observation of colony density and color morphology. This method is difficult to be objective and scientific due to the subjective factors of the observer, and is relatively inefficient.
发明内容Contents of the invention
本发明提供了一种食用菌菌丝长势检测方法，其能够克服现有技术的某种或某些缺陷。The invention provides a method for detecting the growth of edible fungus mycelium, which can overcome some or some defects of the prior art.
根据本发明的一种食用菌菌丝长势检测方法，其在对培养于长方形的栽培盒内的食用菌的菌丝长势进行检测时，以栽培盒作为检测目标；具体包括如下步骤：According to a method for detecting the mycelial growth of edible fungi of the present invention, when detecting the mycelial growth of edible fungi cultured in a rectangular cultivation box, the cultivation box is used as the detection target; specifically, it includes the following steps:
S1、在食用菌生长过程中的不同生长节点处，均对检测目标的灰度图像进行获取；S1. At different growth nodes during the growth process of edible fungi, grayscale images of the detection targets are acquired;
S2、基于当前生长节点的灰度图像Gi+1与前一生长节点的灰度图像Gi，对食用菌在当前生长节点的菌丝长势进行检测。S2. Based on the grayscale image Gi+1 of the current growth node and the grayscale image Gi of the previous growth node, detect the hyphae growth of the edible fungi at the current growth node.
通过上述手段，能够较佳地以图像检测的手段实现对菌丝长势的检测，从而克服常规人工观测所带来的诸如检测结果客观度不足、检测效率较为低下等的缺点。Through the above means, the detection of mycelium growth can be better achieved by means of image detection, thus overcoming the shortcomings of conventional manual observation, such as insufficient objectivity of detection results and low detection efficiency.
作为优选，步骤S2中，Preferably, in step S2,
设置第一灰度阈值Gx和第二灰度阈值Gy，Gx＜Gy；Set the first grayscale threshold Gx and the second grayscale threshold Gy , Gx <Gy ;
获取相应灰度图像中灰度值高于第一灰度阈值Gx的像素点集合，并作为食用菌生长区域；食用菌生长区域中的任一像素点Pxj所对应的灰度值为Gj；Obtain a set of pixels whose grayscale value is higher than the first grayscale threshold Gx in the corresponding grayscale image and serve as the edible fungus growth area; the grayscale value corresponding to any pixel point Pxj in the edible fungi growth area is Gj ;
获取相应灰度图像中灰度值高于第二灰度阈值Gy的像素点集合，并作为优势食用菌生长区域；优势食用菌生长区域中的任一像素点Pxk所对应的灰度值为Gk；Obtain the set of pixels whose grayscale value is higher than the second grayscale threshold Gy in the corresponding grayscale image, and use it as the dominant edible fungus growth area; the grayscale value corresponding to any pixel point Pxk in the dominant edible fungi growth area isGk ;
所述对食用菌在当前生长节点的菌丝长势进行检测，包括对菌丝生长速度Gr1、菌丝密实度Gr2以及菌丝壮实度Gr3的获取；The detection of mycelium growth of edible fungi at the current growth node includes acquisition of mycelium growth rate Gr1 , mycelium density Gr2 and mycelium robustness Gr3 ;
其中，M为灰度图像Gi+1中的食用菌生长区域的像素点总数，N为灰度图像Gi中的食用菌生长区域的像素点总数；Among them, M is the total number of pixels in the edible fungus growth area in the grayscale image Gi+1 , and N is the total number of pixels in the edible fungus growth area in the grayscale image Gi ;
其中，为灰度图像Gi+1中的食用菌生长区域的均化灰度，/>为灰度图像Gi中的食用菌生长区域的均化灰度；in, is the homogenized grayscale of the edible fungus growth area in the grayscale image Gi+1 ,/> is the homogenized grayscale of the edible fungus growth area in the grayscale image Gi ;
其中，为灰度图像Gi+1中的优势食用菌占比，/>为灰度图像Gi中的优势食用菌占比；in, is the proportion of dominant edible fungi in the grayscale image Gi+1 ,/> is the proportion of dominant edible fungi in the grayscale image Gi ;
其中，M0为灰度图像Gi+1中的优势食用菌生长区域的像素点总数，N0为为灰度图像Gi中的优势食用菌生长区域的像素点总数。Among them, M0 is the total number of pixels in the dominant edible fungus growth area in the grayscale image Gi+1 , and N0 is the total number of pixels in the dominant edible fungi growth area in the grayscale image Gi .
基于上述方法，能够基于阈值判定，实现食用菌生长区域以及优势食用菌生长区域的快速寻找定位；而后，即可基于灰度图像中的像素点分布等特性，实现食用菌在当前生长节点的菌丝长势检测，故而简单易行。Based on the above method, the edible fungi growth area and the dominant edible fungus growth area can be quickly found and positioned based on threshold determination; then, based on the pixel distribution and other characteristics in the grayscale image, the edible fungi at the current growth node can be located. The detection of yarn growth is simple and easy.
作为优选，构造权重系数α1、α2和α3，α1+α2+α3＝1；获取菌丝长势指标Q，Q＝α1*Gr1+α2*Gr2+α3*Gr3，“*”为乘积运算。故而能够较佳地获取一个唯一的指标数值以对菌丝长势进行表征，进而便于更为直观地对食用菌的长势变化进行描述。As a preferred method, construct the weight coefficients α1 , α2 and α3 , α1 + α2 + α3 =1; obtain the mycelial growth index Q, Q = α1 *Gr1 + α2 *Gr2 + α3 * Gr3 , "*" is the product operation. Therefore, a unique index value can be better obtained to characterize the growth of mycelium, thereby facilitating a more intuitive description of growth changes of edible fungi.
作为优选，步骤S1中，采用一相机装置对检测目标进行图像采集；对于同一检测目标，以首次对检测目标进行图像采集时检测目标与相机装置间的相对姿态作为基准姿态，在后续的图像采集中控制检测目标与相机装置间的相对姿态与基准姿态保持一致。这使得，能够确保于食用菌生长过程中的每个生长节点处所采集图像均能够具有基本一致的图像特征，从而能够较佳地对抗因拍摄角度不一致所致使的图形畸变等因素而带来的检测精度的降低。Preferably, in step S1, a camera device is used to collect images of the detection target; for the same detection target, the relative posture between the detection target and the camera device when the detection target is imaged for the first time is used as the reference posture. The relative posture between the detection target and the camera device is controlled to be consistent with the reference posture. This ensures that the images collected at each growth node during the growth process of edible fungi can have basically consistent image characteristics, thereby better resisting detection caused by factors such as graphic distortion caused by inconsistent shooting angles. Reduction in accuracy.
作为优选，As a preference,
对于首次对检测目标进行图像采集时所获取的原始图像I0，获取原始图像I0中的4个特征点；该4个特征点分别为P1(xp1，yp1)、P2(xp2，yp2)、P3(xp3，yp3)和P4(xp4，yp4)；其中，特征点P1和P2为一对对角点，特征点P3和P4为另一对对角点；For the original image I0 obtained when the detection target is first collected, four feature points in the original image I0 are obtained; the four feature points are P1 (xp1 , yp1 ), P2 (xp2 , yp2 ), P3 (xp3 , yp3 ) and P4 (xp4 , yp4 ); among them, the feature points P1 and P2 are a pair of diagonal points, and the feature points P3 and P4 are another pair of diagonal points;
对于后续任一次对检测目标进行图像采集时所获取的原始图像获取原始图像/>中的4个对应特征点，该4个对应特征点分别为T1(xT1，yT1)、T2(xT2，yT2)、T3(xT3，yT3)和T4(xT4，yT4)；For the original image obtained during any subsequent image acquisition of the detection target Get original image/>4correspondingfeaturepointsin_____T4 , yT4 );
构建原始图像I0特征矩阵P和原始图像的特征矩阵T，Construct the original image I0 feature matrix P and the original image The characteristic matrix T,
构建原始图像与原始图像I0间的缩放矩阵Z，Build original image The scaling matrix Z between the original image I0 ,
其中，in,
获取原始图像较于原始图像I0间的旋转矩阵R，Get original image Compared with the rotation matrix R between the original image I0 ,
其中，K为相机装置的内参矩阵，Among them, K is the internal parameter matrix of the camera device,
以旋转矩阵R中的u作为航向角、v作为俯仰角，对相机装置的空间姿态进行调整；在完成相机装置的空间姿态调整后，再次对检测目标进行图像采集，并以此时所获取的原始图像Ii+1作为对应生长节点的原始图像。Using u in the rotation matrix R as the heading angle and v as the pitch angle, adjust the spatial attitude of the camera device; after completing the spatial attitude adjustment of the camera device, collect images of the detection target again, and use the obtained The original image Ii+1 is used as the original image of the corresponding growing node.
也即，在后续的每次对检测目标进行图像采集时，能够两次对检测目标进行图像采集，其中第一次所采集的图像能够用于对检测目标与相机装置间的相对姿态进行校正，其中第二次所采集的图像才作为最终菌丝长势判断的依据；故而能够较佳地实现参与菌丝长势检测的每个灰度图像的原始图像均能够具有基本一致的图像特征。That is to say, each subsequent image acquisition of the detection target can be carried out twice, and the image collected for the first time can be used to correct the relative posture between the detection target and the camera device. Only the image collected for the second time is used as the basis for the final judgment of mycelial growth; therefore, it can be better realized that the original image of each grayscale image involved in the hyphal growth detection can have basically consistent image characteristics.
作为优选，As a preference,
基于相同的特征点提取算法获取原始图像I0和原始图像中的所述4个特征点和所述4个对应特征点；Obtain the original image I0 and the original image based on the same feature point extraction algorithm The 4 feature points in and the 4 corresponding feature points;
特征点提取算法具体包括如下步骤：The feature point extraction algorithm specifically includes the following steps:
步骤S11、检测原始图像I0或原始图像的图像边缘；Step S11, detect the original image I0 or the original image the edge of the image;
步骤S12、对位于图像边缘内的每个像素点进行检测，获取位于左上角的第一个特征点P1(xp1，yp1)或T1(xT1，yT1)；Step S12: Detect each pixel located within the edge of the image and obtain the first feature point P1 (xp1 , yp1 ) or T1 (xT1 , yT1 ) located in the upper left corner;
步骤S13、对位于图像边缘内的每个像素点进行检测，获取位于右下角的第二个特征点P2(xp2，yp2)或T2(xT2，yT2)；Step S13: Detect each pixel point located within the edge of the image and obtain the second feature point P2 (xp2 , yp2 ) or T2 (xT2 , yT2 ) located in the lower right corner;
步骤S14，获取位于右上角的第三个特征点P3(xp3，yp3)或T3(xT3，yT3)，以及位于左下角的第四个特征点P4(xp4，yp4)或T4(xT4，yT4)。Step S14, obtain the third feature point P3 (xp3 , yp3 ) or T3 (xT3 , yT3 ) located in the upper right corner, and the fourth feature point P4 (xp4 , y ) located in the lower left corner.p4 ) or T4 (xT4 , yT4 ).
通过上述，能够首先获取第一对对角点，之后依据第一对对角点能够较佳地获取第二对对角点，故而能够较佳地获取两对成正交分布的对角点，该2对对角点在所采集图像上的特征明显且均能够位于视野内，故能够较佳地便于图像处理且能够较佳地表征出图像在行和列方向上的缩放。Through the above, the first pair of diagonal points can be obtained first, and then the second pair of diagonal points can be better obtained based on the first pair of diagonal points. Therefore, two pairs of orthogonally distributed diagonal points can be better obtained. The two pairs of diagonal points have obvious features on the collected image and can both be located within the field of view, so image processing can be better facilitated and the scaling of the image in the row and column directions can be better represented.
作为优选，步骤S12中，采用循环遍历算法，按从上到下逐行遍历、从左到右逐列遍历的方法，对位于图像边缘内的每个像素点进行遍历，进而获取第一个特征点P1(xp1，yp1)或T1(xT1，yT1)。故而能够较佳地实现第一个特征点的获取，且由于循环遍历算法的引入，可较佳避免离散的边缘对特征点识别的干扰。As a preferred method, in step S12, a loop traversal algorithm is used to traverse each pixel located within the edge of the image by row-by-row traversal from top to bottom and column-by-column from left to right, and then obtain the first feature. Point P1 (xp1 , yp1 ) or T1 (xT1 , yT1 ). Therefore, the acquisition of the first feature point can be better achieved, and due to the introduction of the loop traversal algorithm, the interference of discrete edges on feature point recognition can be better avoided.
作为优选，步骤S13中，采用循环遍历算法，按从下到上逐行遍历、从右到左逐列遍历的方法，对位于图像边缘内的每个像素点进行遍历，进而获取第二个特征点P2(xp2，yp2)或T2(xT2，yT2)。同理，故而能够较佳地实现第二个特征点的获取，且由于循环遍历算法的引入，可较佳避免离散的边缘对特征点识别的干扰。As a preferred method, in step S13, a loop traversal algorithm is used to traverse each pixel located within the edge of the image by row-by-row traversal from bottom to top and column-by-column from right to left, and then obtain the second feature. Point P2 (xp2 , yp2 ) or T2 (xT2 , yT2 ). In the same way, the acquisition of the second feature point can be better achieved, and due to the introduction of the loop traversal algorithm, the interference of discrete edges on feature point recognition can be better avoided.
作为优选，步骤S14中，构建特征公式f(d)，Preferably, in step S14, the characteristic formula f(d) is constructed,
f(d)＝(y2-y1)xi-(x2-x1)yi+x2y1-x1y2；f(d)=(y2 -y1 )xi -(x2 -x1 )yi +x2 y1 -x1 y2 ;
以x1＝xp1，y1＝yp1，x2＝xp2，y2＝yp2作为条件代入特征公式f(d)，将f(d)取值最大时所对应的坐标点(xi，yi)作为特征点P3(xp3，yp3)，将f(d)取值最小时所对应的坐标点(xi，yi)作为特征点P4(xp4，yp4)；Substituting x1 =xp1 , y1 =yp1 , x2 =xp2 , y2 =yp2 as conditions into the characteristic formula f(d), and take the coordinate point (x) corresponding to the maximum value of f(d)i , yi ) is used as the feature point P3 (xp3 , yp3 ), and the coordinate point (xi , yi ) corresponding to the minimum value of f(d) is used as the feature point P4 (xp4 , yp4 );
以x1＝xT1，y1＝yT1，x2＝xT2，y2＝yT2作为条件代入特征公式f(d)，将f(d)取值最大时所对应的坐标点(xi，yi)作为特征点T3(xT3，yT3)，将f(d)取值最小时所对应的坐标点(xi，yi)作为特征点T4(xT4，yT4)。Substituting x1 =xT1 , y1 =yT1 , x2 =xT2 , y2 =yT2 into the characteristic formula f(d), and take the coordinate point (x) corresponding to the maximum value of f(d)i , yi ) as the feature point T3 (xT3 , yT3 ), and the coordinate point (xi , yi ) corresponding to the minimum value of f(d) is used as the feature point T4 (xT4 , yT4 ).
通过上述，能够较佳地基于距离公式实现对第三和第四特征点的求取。Through the above, the third and fourth feature points can be obtained preferably based on the distance formula.
此外本发明还提供了一种食用菌菌丝长势检测装置，其包括相机装置及处理装置，相机装置用于实现对检测目标的图像采集，处理装置用于基于相机装置所采集的图像对食用菌在当前生长节点的菌丝长势进行检测。故而能够较佳地实现对菌丝长势的自动检测。In addition, the present invention also provides an edible fungus mycelium growth detection device, which includes a camera device and a processing device. The camera device is used to collect images of the detection target, and the processing device is used to detect edible fungi based on the images collected by the camera device. The hyphae growth at the current growth node is detected. Therefore, the automatic detection of mycelium growth can be better realized.
附图说明Description of the drawings
图1为实施例1中的检测方法的示意图。Figure 1 is a schematic diagram of the detection method in Example 1.
具体实施方式Detailed ways
为进一步了解本发明的内容，结合附图和实施例对本发明作详细描述。应当理解的是，实施例仅仅是对本发明进行解释而并非限定。In order to further understand the content of the present invention, the present invention will be described in detail with reference to the accompanying drawings and embodiments. It should be understood that the embodiments are only for explanation of the present invention but not for limitation.
实施例1Example 1
考虑到菌丝的长势检测中，能够基于菌落的生长尺寸实现对菌丝生长速度的检测；并考虑到大部分食用菌的菌丝在生长前期大多数都是白色的，故能够基于颜色的深浅实现对菌丝的密实度及壮实度的检测，也即菌丝越密实越白、菌丝越壮实越白。Considering that in the detection of mycelium growth, the mycelium growth rate can be detected based on the growth size of the colony; and considering that most of the mycelium of edible fungi are white in the early stage of growth, it can be based on the depth of color. Realize the detection of the density and robustness of the hyphae, that is, the denser the hyphae, the whiter, and the stronger the hyphae, the whiter.
故本实施例基于图像检测的手段，提供了一种食用菌菌丝长势检测方法。见于图1，在采用本实施例的方法在对培养于长方形的栽培盒内的食用菌的菌丝长势进行检测时，以栽培盒作为检测目标(此处以栽培有食用菌的栽培盒整体作为检测目标，能够利用栽培盒呈长方形的形状特征且在图像中边缘轮廓清晰的特性，较佳实现下述中的对拍摄角度的校正)；具体包括如下步骤：Therefore, this embodiment provides a method for detecting the growth of edible fungus hyphae based on image detection. As shown in Figure 1, when using the method of this embodiment to detect the mycelial growth of edible fungi cultured in a rectangular cultivation box, the cultivation box is used as the detection target (here, the entire cultivation box cultivated with edible fungi is used as the detection target). The goal is to use the rectangular shape characteristics of the cultivation box and the clear edge outline in the image to best achieve the following correction of the shooting angle); the specific steps include the following:
S1、在食用菌生长过程中的不同生长节点处，均对检测目标的灰度图像进行获取；S1. At different growth nodes during the growth process of edible fungi, grayscale images of the detection targets are acquired;
S2、基于当前生长节点的灰度图像Gi+1与前一生长节点的灰度图像Gi，对食用菌在当前生长节点的菌丝长势进行检测。S2. Based on the grayscale image Gi+1 of the current growth node and the grayscale image Gi of the previous growth node, detect the hyphae growth of the edible fungi at the current growth node.
通过上述手段，能够较佳地以图像检测的手段实现对菌丝长势的检测，从而克服常规人工观测所带来的诸如检测结果客观度不足、检测效率较为低下等的缺点。Through the above means, the detection of mycelium growth can be better achieved by means of image detection, thus overcoming the shortcomings of conventional manual observation, such as insufficient objectivity of detection results and low detection efficiency.
见于下文，对检测目标的灰度图像的获取，能够首先以相机装置采集RGB图像作为原始图像，而后通过处理装置将每个像素点的R、G、B值转换为灰度值即可实现。关于R、G、B值转换为灰度值的转换方法或公式，采用现有常规的转换公式即可，本实施例中不予赘述。As shown below, the grayscale image of the detection target can be obtained by first collecting the RGB image as the original image with the camera device, and then converting the R, G, and B values of each pixel into grayscale values through the processing device. Regarding the conversion method or formula for converting R, G, and B values into grayscale values, existing conventional conversion formulas can be used, and will not be described in detail in this embodiment.
在步骤S2中，In step S2,
能够设置第一灰度阈值Gx和第二灰度阈值Gy，Gx＜Gy；The first grayscale threshold Gx and the second grayscale threshold Gy can be set, Gx <Gy ;
获取相应灰度图像中灰度值高于第一灰度阈值Gx的像素点集合，并作为食用菌生长区域；食用菌生长区域中的任一像素点Pxj所对应的灰度值为Gj；Obtain a set of pixels whose grayscale value is higher than the first grayscale threshold Gx in the corresponding grayscale image and serve as the edible fungus growth area; the grayscale value corresponding to any pixel point Pxj in the edible fungi growth area is Gj ;
获取相应灰度图像中灰度值高于第二灰度阈值Gy的像素点集合，并作为优势食用菌生长区域；优势食用菌生长区域中的任一像素点Pxk所对应的灰度值为Gk；Obtain the set of pixels whose grayscale value is higher than the second grayscale threshold Gy in the corresponding grayscale image and serve as the dominant edible fungus growth area; the grayscale value corresponding to any pixel point Pxk in the dominant edible fungi growth area isGk ;
所述对食用菌在当前生长节点的菌丝长势进行检测，包括对菌丝生长速度Gr1、菌丝密实度Gr2以及菌丝壮实度Gr3的获取；The detection of mycelium growth of edible fungi at the current growth node includes acquisition of mycelium growth rate Gr1 , mycelium density Gr2 and mycelium robustness Gr3 ;
其中，M为灰度图像Gi+1中的食用菌生长区域的像素点总数，N为灰度图像Gi中的食用菌生长区域的像素点总数；Among them, M is the total number of pixels in the edible fungus growth area in the grayscale image Gi+1 , and N is the total number of pixels in the edible fungus growth area in the grayscale image Gi ;
其中，为灰度图像Gi+1中的食用菌生长区域的均化灰度，/>为灰度图像Gi中的食用菌生长区域的均化灰度；in, is the homogenized grayscale of the edible fungus growth area in the grayscale image Gi+1 ,/> is the homogenized grayscale of the edible fungus growth area in the grayscale image Gi ;
其中，为灰度图像Gi+1中的优势食用菌占比，/>为灰度图像Gi中的优势食用菌占比；in, is the proportion of dominant edible fungi in the grayscale image Gi+1 ,/> is the proportion of dominant edible fungi in the grayscale image Gi ;
其中，M0为灰度图像Gi+1中的优势食用菌生长区域的像素点总数，N0为为灰度图像Gi中的优势食用菌生长区域的像素点总数。Among them, M0 is the total number of pixels in the dominant edible fungus growth area in the grayscale image Gi+1 , and N0 is the total number of pixels in the dominant edible fungi growth area in the grayscale image Gi .
基于上述方法，能够基于阈值判定，实现食用菌生长区域以及优势食用菌生长区域的快速寻找定位；而后，即可基于灰度图像中的像素点分布等特性，实现食用菌在当前生长节点的菌丝长势检测，故而简单易行。Based on the above method, the edible fungi growth area and the dominant edible fungus growth area can be quickly found and positioned based on threshold determination; then, based on the pixel distribution and other characteristics in the grayscale image, the edible fungi at the current growth node can be located. The detection of yarn growth is simple and easy.
可以理解的是，食用菌的培养基或培养土均为深色(灰度值更小)，而食用菌的菌丝为白色(灰度值更大)，故能够较佳地利用阈值判定的方式实现菌丝所对应像素点的检出。通过筛选灰度值高于第一灰度阈值Gx的像素点集合，即可较佳地筛选出属于食用菌菌丝的像素点；通过筛选灰度值高于第二灰度阈值Gy的像素点集合，即可较佳地筛选出属于优势食用菌菌丝的像素点；可以理解的是优势食用菌菌丝具有更白的色度。It can be understood that the culture medium or culture soil of edible fungi is dark (the gray value is smaller), while the mycelium of the edible fungi is white (the gray value is larger), so the threshold value judgment can be better used. This method realizes the detection of pixels corresponding to hyphae. By screening the set of pixelswhose grayscale value is higher than the first grayscale thresholdG By collecting pixel points, the pixel points belonging to the dominant edible fungus hyphae can be better screened out; it can be understood that the dominant edible fungus hyphae have a whiter color.
并且，可以理解的是，第一灰度阈值Gx和第二灰度阈值Gy能够依照经验数据进行设置。Moreover, it can be understood that the first grayscale threshold Gx and the second grayscale threshold Gy can be set according to empirical data.
同时，本实施例中还能够构造权重系数α1、α2和α3，α1+α2+α3＝1；获取菌丝长势指标Q，Q＝α1*Gr1+α2*Gr2+α3*Gr3，“*”为乘积运算。故而能够较佳地获取一个唯一的指标数值以对菌丝长势进行表征，进而便于更为直观地对食用菌的长势变化进行描述。At the same time, in this embodiment, the weight coefficients α1 , α2 and α3 can also be constructed, α1 + α2 + α3 = 1; the mycelium growth index Q is obtained, Q = α1 *Gr1 + α2 *Gr2 +α3 *Gr3 , "*" is the product operation. Therefore, a unique index value can be better obtained to characterize the growth of mycelium, thereby facilitating a more intuitive description of growth changes of edible fungi.
其中，权重系数α1、α2和α3能够依据经验进行获取。Among them, the weight coefficients α1 , α2 and α3 can be obtained based on experience.
此外，在步骤S1中，还能够采用一相机装置对检测目标进行图像采集；对于同一检测目标，以首次对检测目标进行图像采集时检测目标与相机装置间的相对姿态作为基准姿态，在后续的图像采集中控制检测目标与相机装置间的相对姿态与基准姿态保持一致。这使得，能够确保于食用菌生长过程中的每个生长节点处所采集图像均能够具有基本一致的图像特征，从而能够较佳地对抗因拍摄角度不一致所致使的图形畸变等因素而带来的检测精度的降低。In addition, in step S1, a camera device can also be used to collect images of the detection target; for the same detection target, the relative posture between the detection target and the camera device when the detection target is imaged for the first time is used as the reference posture. During image acquisition, the relative posture between the detection target and the camera device is controlled to be consistent with the reference posture. This ensures that the images collected at each growth node during the growth process of edible fungi can have basically consistent image characteristics, thereby better resisting detection caused by factors such as graphic distortion caused by inconsistent shooting angles. Reduction in accuracy.
其中：in:
对于首次对检测目标进行图像采集时所获取的原始图像I0，获取原始图像I0中的4个特征点；该4个特征点分别为P1(xp1，yp1)、P2(xp2，yp2)、P3(xp3，yp3)和P4(xp4，yp4)；其中，特征点P1和P2为一对对角点，特征点P3和P4为另一对对角点；For the original image I0 obtained when the detection target is first collected, four feature points in the original image I0 are obtained; the four feature points are P1 (xp1 , yp1 ), P2 (xp2 , yp2 ), P3 (xp3 , yp3 ) and P4 (xp4 , yp4 ); among them, the feature points P1 and P2 are a pair of diagonal points, and the feature points P3 and P4 are another pair of diagonal points;
对于后续任一次对检测目标进行图像采集时所获取的原始图像获取原始图像/>中的4个对应特征点，该4个对应特征点分别为T1(xT1，yT1)、T2(xT2，yT2)、T3(xT3，yT3)和T4(xT4，yT4)；For the original image obtained during any subsequent image acquisition of the detection target Get original image/>4correspondingfeaturepointsin_____T4 , yT4 );
构建原始图像I0特征矩阵P和原始图像的特征矩阵T，Construct the original image I0 feature matrix P and the original image The characteristic matrix T,
构建原始图像与原始图像I0间的缩放矩阵Z，Build original image The scaling matrix Z between the original image I0 ,
其中，in,
获取原始图像较于原始图像I0间的旋转矩阵R，Get original image Compared with the rotation matrix R between the original image I0 ,
其中，K为相机装置的内参矩阵，Among them, K is the internal parameter matrix of the camera device,
以旋转矩阵R中的u作为航向角、v作为俯仰角，对相机装置的空间姿态进行调整；在完成相机装置的空间姿态调整后，再次对检测目标进行图像采集，并以此时所获取的原始图像Ii+1作为对应生长节点的原始图像。Using u in the rotation matrix R as the heading angle and v as the pitch angle, adjust the spatial attitude of the camera device; after completing the spatial attitude adjustment of the camera device, collect images of the detection target again, and use the obtained The original image Ii+1 is used as the original image of the corresponding growing node.
也即，在后续的每次对检测目标进行图像采集时，能够两次对检测目标进行图像采集，其中第一次所采集的图像能够用于对检测目标与相机装置间的相对姿态进行校正，其中第二次所采集的图像才作为最终菌丝长势判断的依据；故而能够较佳地实现参与菌丝长势检测的每个灰度图像的原始图像均能够具有基本一致的图像特征。That is to say, each subsequent image acquisition of the detection target can be carried out twice, and the image collected for the first time can be used to correct the relative posture between the detection target and the camera device. Only the image collected for the second time is used as the basis for the final judgment of mycelial growth; therefore, it can be better realized that the original image of each grayscale image involved in the hyphal growth detection can have basically consistent image characteristics.
可以理解的是，特征点P1(xp1，yp1)和对应特征点T1(xT1，yT1)分别为对应图像中的最左和最上的像素点，特征点P2(xp2，yp2)和对应特征点T2(xT2，yT2)为对应图像中的最右和最下的像素点。这种构造与栽培盒的长方形的特性相关，也就是说能够以栽培盒的4个顶角所在位置实现对检测目标与相机装置间的相对姿态的校正。It can be understood that the feature point P1 (xp1 , yp1 ) and the corresponding feature point T1 (xT1 , yT1 ) are the leftmost and top pixels in the corresponding image respectively, and the feature point P2 (xp2 , yp2 ) and the corresponding feature point T2 (xT2 , yT2 ) are the rightmost and bottom pixels in the corresponding image. This structure is related to the rectangular characteristics of the cultivation box, which means that the relative posture between the detection target and the camera device can be corrected based on the positions of the four vertex corners of the cultivation box.
可以理解的是，本实施例中的相机装置能够通过一云台设置于固定高度处，通过以旋转矩阵R中的u作为航向角、v作为俯仰角作为云台的控制指令，即可实现相机装置的拍摄角度的自动调整。It can be understood that the camera device in this embodiment can be set at a fixed height through a pan/tilt. By using u in the rotation matrix R as the heading angle and v as the pitch angle as the control instructions of the pan/tilt, the camera can be realized Automatic adjustment of the device's shooting angle.
其中，本实施例中的航向角u是指监控相机的拍摄轴线与水平方向的偏移量，本实施例中的俯仰角v是指监控相机的拍摄轴线与竖直方向的偏移量。The heading angle u in this embodiment refers to the offset between the shooting axis of the surveillance camera and the horizontal direction, and the pitch angle v in this embodiment refers to the offset between the shooting axis of the surveillance camera and the vertical direction.
本实施例中，能够基于相同的特征点提取算法获取原始图像I0和原始图像中的所述4个特征点和所述4个对应特征点；In this embodiment, the original image I0 and the original image can be obtained based on the same feature point extraction algorithm The 4 feature points in and the 4 corresponding feature points;
特征点提取算法具体包括如下步骤：The feature point extraction algorithm specifically includes the following steps:
步骤S11、检测原始图像I0或原始图像的图像边缘；Step S11, detect the original image I0 or the original image the edge of the image;
步骤S12、对位于图像边缘内的每个像素点进行检测，获取位于左上角的第一个特征点P1(xp1，yp1)或T1(xT1，yT1)；Step S12: Detect each pixel located within the edge of the image and obtain the first feature point P1 (xp1 , yp1 ) or T1 (xT1 , yT1 ) located in the upper left corner;
步骤S13、对位于图像边缘内的每个像素点进行检测，获取位于右下角的第二个特征点P2(xp2，yp2)或T2(xT2，yT2)；Step S13: Detect each pixel point located within the edge of the image and obtain the second feature point P2 (xp2 , yp2 ) or T2 (xT2 , yT2 ) located in the lower right corner;
步骤S14，获取位于右上角的第三个特征点P3(xp3，yp3)或T3(xT3，yT3)，以及位于左下角的第四个特征点P4(xp4，yp4)或T4(xT4，yT4)。Step S14, obtain the third feature point P3 (xp3 , yp3 ) or T3 (xT3 , yT3 ) located in the upper right corner, and the fourth feature point P4 (xp4 , y ) located in the lower left corner.p4 ) or T4 (xT4 , yT4 ).
通过上述，能够首先获取第一对对角点，之后依据第一对对角点能够较佳地获取第二对对角点，故而能够较佳地获取两对成正交分布的对角点，该2对对角点在所采集图像上的特征明显且均能够位于视野内，故能够较佳地便于图像处理且能够较佳地表征出图像在行和列方向上的缩放。Through the above, the first pair of diagonal points can be obtained first, and then the second pair of diagonal points can be better obtained based on the first pair of diagonal points. Therefore, two pairs of orthogonally distributed diagonal points can be better obtained. The two pairs of diagonal points have obvious features on the collected image and can both be located within the field of view, so image processing can be better facilitated and the scaling of the image in the row and column directions can be better represented.
其中，本实施例的步骤S11中，能够基于常规的如Canny算子实现图像边缘的检测，故而能够较佳地实现对栽培盒的边缘识别。Among them, in step S11 of this embodiment, the image edge can be detected based on a conventional Canny operator, so that the edge recognition of the cultivation box can be better realized.
在本实施例的步骤S12中，能够采用例如循环遍历算法，按从上到下逐行遍历、从左到右逐列遍历的方法，对位于图像边缘内的每个像素点进行遍历，进而获取第一个特征点P1(xp1，yp1)或T1(xT1，yT1)。故而能够较佳地实现第一个特征点的获取，且由于循环遍历算法的引入，可较佳避免离散的边缘对特征点识别的干扰。In step S12 of this embodiment, for example, a loop traversal algorithm can be used to traverse each pixel located within the edge of the image by row-by-row traversal from top to bottom and column-by-column from left to right, and then obtain The first feature point P1 (xp1 , yp1 ) or T1 (xT1 , yT1 ). Therefore, the acquisition of the first feature point can be better achieved, and due to the introduction of the loop traversal algorithm, the interference of discrete edges on feature point recognition can be better avoided.
在本实施例的步骤S13中，能够采用循环遍历算法，按从下到上逐行遍历、从右到左逐列遍历的方法，对位于图像边缘内的每个像素点进行遍历，进而获取第二个特征点P2(xp2，yp2)或T2(xT2，yT2)。同理，故而能够较佳地实现第二个特征点的获取，且由于循环遍历算法的引入，可较佳避免离散的边缘对特征点识别的干扰。In step S13 of this embodiment, a loop traversal algorithm can be used to traverse each pixel located within the edge of the image by row-by-row traversal from bottom to top and column-by-column from right to left, and then obtain the third Two feature points P2 (xp2 , yp2 ) or T2 (xT2 , yT2 ). In the same way, the acquisition of the second feature point can be better achieved, and due to the introduction of the loop traversal algorithm, the interference of discrete edges on feature point recognition can be better avoided.
在本实施例的步骤S14中，能够构建特征公式f(d)，In step S14 of this embodiment, the characteristic formula f(d) can be constructed,
f(d)＝(y2-y1)xi-(x2-x1)yi+x2y1-x1y2；f(d)=(y2 -y1 )xi -(x2 -x1 )yi +x2 y1 -x1 y2 ;
以x1＝xp1，y1＝yp1，x2＝xp2，y2＝yp2作为条件代入特征公式f(d)，将f(d)取值最大时所对应的坐标点(xi，yi)作为特征点P3(xp3，yp3)，将f(d)取值最小时所对应的坐标点(xi，yi)作为特征点P4(xp4，yp4)；Substituting x1 =xp1 , y1 =yp1 , x2 =xp2 , y2 =yp2 as conditions into the characteristic formula f(d), and take the coordinate point (x) corresponding to the maximum value of f(d)i , yi ) is used as the feature point P3 (xp3 , yp3 ), and the coordinate point (xi , yi ) corresponding to the minimum value of f(d) is used as the feature point P4 (xp4 , yp4 );
以x1＝xT1，y1＝yT1，x2＝xT2，y2＝yT2作为条件代入特征公式f(d)，将f(d)取值最大时所对应的坐标点(xi，yi)作为特征点T3(xT3，yT3)，将f(d)取值最小时所对应的坐标点(xi，yi)作为特征点T4(xT4，yT4)。Substituting x1 =xT1 , y1 =yT1 , x2 =xT2 , y2 =yT2 into the characteristic formula f(d), and take the coordinate point (x) corresponding to the maximum value of f(d)i , yi ) as the feature point T3 (xT3 , yT3 ), and the coordinate point (xi , yi ) corresponding to the minimum value of f(d) is used as the feature point T4 (xT4 , yT4 ).
通过上述，能够较佳地基于距离公式实现对第三和第四特征点的求取。Through the above, the third and fourth feature points can be obtained preferably based on the distance formula.
此处进行进一步解释的是，在第一和第二特征点确定后，其两点连线的直线L的方程能够表达为：What is further explained here is that after the first and second feature points are determined, the equation of the straight line L connecting the two points can be expressed as:
(y2-y1)x-(x2-x1)y+x2y1-x1y2＝0。(y2 -y1 )x-(x2 -x1 )y+x2 y1 -x1 y2 =0.
图像中任一像素点(xi，yi)到直线L的垂直距离能够表达为：The vertical distance from any pixel point (xi , yi ) in the image to the straight line L can be expressed as:
上述距离公式中的分母为常数，故本实施例中能够简化获取特征公式f(d)，故而能够较佳地降低计算量。通过对特征公式f(d)的最大值和最小值的求取，即可较佳地获取第三和第四特征点。The denominator in the above distance formula is a constant, so in this embodiment, it is possible to simplify the acquisition of the characteristic formula f(d), and thus the amount of calculation can be reduced better. By finding the maximum and minimum values of the characteristic formula f(d), the third and fourth characteristic points can be better obtained.
此外，为了实现本实施例的方法，本实施例还提供了一种食用菌菌丝长势检测装置，其包括相机装置及处理装置，相机装置用于实现对检测目标的图像采集，处理装置用于基于相机装置所采集的图像对食用菌在当前生长节点的菌丝长势进行检测。故而能够较佳地实现对菌丝长势的自动检测。In addition, in order to implement the method of this embodiment, this embodiment also provides an edible fungus mycelium growth detection device, which includes a camera device and a processing device. The camera device is used to collect images of the detection target, and the processing device is used to collect images of the detection target. Based on the images collected by the camera device, the mycelium growth of the edible fungi at the current growth node is detected. Therefore, the automatic detection of mycelium growth can be better realized.
以上示意性的对本发明及其实施方式进行了描述，该描述没有限制性，附图中所示的也只是本发明的实施方式之一，实际的结构并不局限于此。所以，如果本领域的普通技术人员受其启示，在不脱离本发明创造宗旨的情况下，不经创造性的设计出与该技术方案相似的结构方式及实施例，均应属于本发明的保护范围。The present invention and its embodiments are schematically described above. This description is not limiting. What is shown in the drawings is only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person of ordinary skill in the art is inspired by the invention and without departing from the spirit of the invention, can devise structural methods and embodiments similar to the technical solution without inventiveness, they shall all fall within the protection scope of the invention. .
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311451695.0ACN117495802B (en) | 2023-11-02 | 2023-11-02 | A method and device for detecting the growth of edible fungi mycelium |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311451695.0ACN117495802B (en) | 2023-11-02 | 2023-11-02 | A method and device for detecting the growth of edible fungi mycelium |
| Publication Number | Publication Date |
|---|---|
| CN117495802Atrue CN117495802A (en) | 2024-02-02 |
| CN117495802B CN117495802B (en) | 2025-04-15 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311451695.0AActiveCN117495802B (en) | 2023-11-02 | 2023-11-02 | A method and device for detecting the growth of edible fungi mycelium |
| Country | Link |
|---|---|
| CN (1) | CN117495802B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120107196A (en)* | 2025-02-17 | 2025-06-06 | 靖州康源苓业科技股份有限公司 | A method and device for detecting the activity of Poria cocos strains |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102313518A (en)* | 2010-07-02 | 2012-01-11 | 上海市农业科学院 | Quantitative acquisition method for growth of microorganism hypha |
| CN107164452A (en)* | 2017-06-06 | 2017-09-15 | 武汉市农业科学技术研究院蔬菜科学研究所 | A kind of detection method of morchella fungal bacterial strain cell age |
| CN107248159A (en)* | 2017-08-04 | 2017-10-13 | 河海大学常州校区 | A kind of metal works defect inspection method based on binocular vision |
| CN108564124A (en)* | 2018-04-13 | 2018-09-21 | 山东农业大学 | A kind of magnaporthe grisea spore microimage detection recognition methods based on support vector machines |
| CN110610506A (en)* | 2019-09-17 | 2019-12-24 | 北京中环易达设施园艺科技有限公司 | Image processing technology-based agaricus blazei murill fruiting body growth parameter detection method |
| CA3115544A1 (en)* | 2018-10-24 | 2020-04-30 | Philip Ross | Monokaryon mycelial material and related method of production |
| CN113455289A (en)* | 2021-08-05 | 2021-10-01 | 河南省农业科学院植物营养与资源环境研究所 | Method for rapidly detecting fruiting performance of edible mushroom tissue isolated strain |
| CN116948840A (en)* | 2023-06-13 | 2023-10-27 | 山东省农业科学院 | Hypsizygus marmoreus HL-5 and cultivation method thereof |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102313518A (en)* | 2010-07-02 | 2012-01-11 | 上海市农业科学院 | Quantitative acquisition method for growth of microorganism hypha |
| CN107164452A (en)* | 2017-06-06 | 2017-09-15 | 武汉市农业科学技术研究院蔬菜科学研究所 | A kind of detection method of morchella fungal bacterial strain cell age |
| CN107248159A (en)* | 2017-08-04 | 2017-10-13 | 河海大学常州校区 | A kind of metal works defect inspection method based on binocular vision |
| CN108564124A (en)* | 2018-04-13 | 2018-09-21 | 山东农业大学 | A kind of magnaporthe grisea spore microimage detection recognition methods based on support vector machines |
| CA3115544A1 (en)* | 2018-10-24 | 2020-04-30 | Philip Ross | Monokaryon mycelial material and related method of production |
| CN110610506A (en)* | 2019-09-17 | 2019-12-24 | 北京中环易达设施园艺科技有限公司 | Image processing technology-based agaricus blazei murill fruiting body growth parameter detection method |
| CN113455289A (en)* | 2021-08-05 | 2021-10-01 | 河南省农业科学院植物营养与资源环境研究所 | Method for rapidly detecting fruiting performance of edible mushroom tissue isolated strain |
| CN116948840A (en)* | 2023-06-13 | 2023-10-27 | 山东省农业科学院 | Hypsizygus marmoreus HL-5 and cultivation method thereof |
| Title |
|---|
| 张丹;朱爱莲;赵静;周敏;张引芳;尚晓冬;王瑞娟;谭琦;: "采用木屑培养基传代的金针菇菌种稳定性", 食用菌学报, no. 02, 15 June 2018 (2018-06-15)* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120107196A (en)* | 2025-02-17 | 2025-06-06 | 靖州康源苓业科技股份有限公司 | A method and device for detecting the activity of Poria cocos strains |
| Publication number | Publication date |
|---|---|
| CN117495802B (en) | 2025-04-15 |
| Publication | Publication Date | Title |
|---|---|---|
| Boogaard et al. | Robust node detection and tracking in fruit-vegetable crops using deep learning and multi-view imaging | |
| CN106468543B (en) | Blade area measuring method based on image processing | |
| CN112581904B (en) | A Moiré Compensation Method for Brightness Grayscale Image of OLED Screen | |
| CN112395984B (en) | Method for detecting seedling guide line of unmanned agricultural machine | |
| CN109741393B (en) | Diameter measurement and center point location method of Agaricus bisporus | |
| CN110070571B (en) | Phyllostachys pubescens morphological parameter detection method based on depth camera | |
| CN110455201B (en) | Stalk crop height measuring method based on machine vision | |
| CN114119437B (en) | GMS-based image stitching method for improving distortion of moving object | |
| CN107392956A (en) | Crop root Phenotypic examination method and apparatus | |
| CN117495802A (en) | A method and device for detecting the growth of edible fungus hyphae | |
| CN111882549B (en) | Automatic detection and identification method and detection and identification system for grayish green tiny foreign fibers | |
| CN111724354B (en) | Image processing-based method for measuring wheat ear length and wheat ear number of multiple wheat plants | |
| CN104200457A (en) | Wide-angle camera shooting based discrete type canopy leaf area index detection system and method | |
| CN118968117A (en) | Screening and grading system and method for fresh cut peony flowers based on image recognition | |
| CN114838664A (en) | In-situ pileus size measuring method based on black-skin termitomyces albuminosus | |
| CN114037832A (en) | Method and system for extracting typical region in satellite remote sensing image | |
| CN114387343A (en) | Method for detecting picking position of mushroom stem of clustered oyster mushroom | |
| CN109166127B (en) | A wearable plant phenotype sensing system | |
| TWI383690B (en) | Method for image processing | |
| CN111678441A (en) | A device and method for underwater fish shape measurement based on deep learning | |
| CN115330868A (en) | Grape picking method based on deep learning and depth information fusion | |
| CN112070723B (en) | Automatic identification method for plain woven fabric density | |
| CN114299144A (en) | An accurate positioning method of rice based on machine vision | |
| CN105574853A (en) | Method and system for calculating number of wheat grains based on image identification | |
| CN114838665B (en) | A kind of in-situ measurement method of size based on Gallicardia nigra |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |