Single-plant standing timber volume nondestructive measurement method suitable for small and medium-sized arbors
One, the technical field
The invention relates to a method for calculating the volume of a single standing tree, which is mainly a method for calculating the volume of the single standing tree by using a smart phone.
Second, technical background
The forest ecosystem is vital to the survival of animals and humans, not only provides services in the aspects of carbon cycle, water and soil conservation, climate regulation, biodiversity and the like, but also provides food, wood and energy for the humans. The forest accumulation is the total volume of all standing timbers in the forest, and the forest accumulation is not only an important factor for evaluating the forest resource state and the forest management level, but also an important index for evaluating the capability of the forest for reducing the climate change risk, so the forest accumulation is one of the main factors for forest investigation. The measurement of the volume of the standing timber is the basis for accurately evaluating the forest accumulation amount and is also the key point of tree research for a long time.
The method mainly measures the volume of standing trees by a trunk analysis method in the early stage, analyzes the trunk after sample trees are felled, and obtains a modeling sample required for compiling the volume table. Accordingly, some researchers have proposed establishing non-destructive wood volume estimation methods such as experimental shape number, shape point, orthomorphism, hope-height, and the like. However, these methods are limited in their use due to the disadvantages of cumbersome operation and low measurement accuracy. Therefore, scholars at home and abroad hope to establish a nondestructive testing method capable of quickly and accurately measuring the volume of wood. In recent years, ground observation technology, remote sensing technology, photogrammetry technology, 3D laser scanning technology and the like provide effective solutions for forest resource inventory, and have been used for extracting various forestry attributes, but three-dimensional laser scanners, ground laser radars and the like are more suitable for large-scale forest resource investigation due to high cost and difficult data processing. In China, multifunctional comprehensive tree testers such as a handheld electronic tree-measuring gun, a CCD super-station instrument, a forest intelligent surveying and mapping calculator, a miniature super-station instrument, a laser photography tree-measuring instrument and the like appear in succession, and a relatively complete forest ground investigation technical system is formed. Although the above-described techniques may address the problem of measuring standing timber volume, these associated measuring instruments require high cost or specialized training. The existing forestry observation equipment still has the defects of high cost, low measurement precision, low integration level, poor portability, single function and the like. The progress of computer vision helps researchers develop methods which are very suitable for forest investigation and have the characteristics of high precision and low price. The feasibility of this method has been verified by computer vision and ground observation techniques to obtain standing timber volume, but the equipment using this technique is usually a stand-alone digital camera, requires the photographic film to be imported into computer software for processing, and cannot be resolved on site to obtain measurements. In the existing forest investigation method, the problems of low measurement precision, complex operation, heavy equipment or high cost and the like still exist particularly in the aspect of calculating the volume of standing timber. These factors not only limit the popularization and application of the equipment in forest investigation, but also restrict the computerization of modern forestry.
The invention designs and develops a forest electronic tree measuring instrument, which highly integrates a smart phone with Bluetooth data transmission function and a laser range finder, integrates the functions of measuring angles and distances by combining forest survey software which is developed based on Java language and runs in the smart phone, and can measure the volume of standing trees in real time, quickly and accurately. The forest electronic tree measuring instrument has the advantages of being cheaper than a 3D laser scanner, convenient to carry, simple to operate and free of post processing. Compared with the traditional method, the invention reduces the field working time and workload, can be independently operated by one person without the assistance of extra personnel, and realizes the real non-contact real-time measurement. In addition, the total station and the forest electronic tree measuring instrument are respectively used for measuring the volume of the standing timber, the obtained volume data of the standing timber are compared and analyzed, and the feasibility of accurately measuring the volume of the standing timber by using the forest electronic tree measuring instrument is verified.
Third, the invention
Aiming at the defects and shortcomings of the existing technology for calculating the volume of the single standing tree, the traditional method for measuring the volume of the single standing tree is supplemented. The invention provides a method for obtaining the volume of a single standing tree with higher precision and lower cost without damaging the tree. The invention aims to provide a nondestructive measurement method for the volume of single standing tree used for small and medium-sized trees (5 m-30 m).
The main invention content is as follows:
1. a method for calculating the volume of a single standing tree by using a Forest Survey Intelligent tree measurer (FSID) is characterized in that: the method comprises the steps of obtaining photos of standing trees by using an independently developed FSID (free space identification) in combination with a double-point observation method, carrying out 'standing tree distinguishing and integrating' by using an improved felling tree distinguishing and integrating algorithm, carrying out standing tree trunk analysis and trunk data calculation, and finally carrying out standing tree volume calculation to obtain the volume of a single standing tree.
2. The FSID consists of hardware and software. In the aspect of hardware, an Android 8.0 operating system is mounted, and various sensors such as an angle sensor, a camera and Bluetooth are equipped. The instrument has the advantages that the core module is a mobile phone, the total weight is less than 2kg, the carrying is convenient, the operation is simple, and the instrument is highly integrated with the mobile phone, the laser range finder and the photographic platform. The instrument can perform software operation and real-time data processing, controls interaction of other hardware modules, and provides high-precision data for the mobile phone by adopting BLE4.0 low-power Bluetooth communication technology after ranging of the laser range finder. The photographing platform is a tripod, integrates hardware connection, is used for placing and fixing a mobile phone, can rotate 360 degrees, and can be adjusted at will in height relative to the ground. In terms of software, on an Android Studio 3.3 development platform, the FSID uses Java programming language to develop forest survey software, and uses Java Native Interface (JNI) technology to call an Open Source Computer Vision Library (OpenCV) interface to process a standing tree photo. The software can measure the angle, the distance and the volume of the standing timber in real time and carry out photo splicing on the shot standing timber. The user can adjust the instrument according to the actual measurement environment and the measurement work requirement, and angle data, distance data and pictures are processed. All measurement data may be stored or exported (. csv) in SQLite database format.
3. The double-point observation method is characterized in that the following measurements are respectively carried out at two observation points: (1) and erecting a forest electronic forest measuring instrument. And (4) opening forest measurement software, and measuring the horizontal distance from the observation point to the trunk by using a laser range finder. The horizontal distance is transmitted to the handset via bluetooth. Then, by using the rotary platform, the cross-hair at the center of the camera is aligned with a point at the top of the tree and a point at the bottom of the tree (a part of the trunk contacting with the ground) to respectively obtain the angle between the observation field and the top of the tree and the angle between the observation field and the bottom of the tree. (2) Keeping the main optical axis of the camera parallel to the ground, taking a standing tree photo, wherein measurement parameters comprise an included angle alpha between an observation station and the top of a tree, an included angle beta between the observation station and the bottom of the tree, a horizontal distance L between an instrument and the station, and a plurality of intersection points of a calibration line and a trunk in the standing tree photo. The photographs were processed when the observation conditions were difficult.
4. The photo processing means that when the observation place is close to the standing tree or the standing tree to be measured is high, the whole information of the standing tree cannot be included by photographing. Therefore, the FSID photographing mode is changed, and a plurality of photographs are taken of the trees having the overlapping area. The software developed by the method corrects the acquired photos one by one in a horizontal tilting mode. And splicing the multiple photos of the overlapped parts together to form a photo containing the complete information of the upright wood. In case the tree volume cannot be measured from one photo, two or three photos with overlapping regions are merged.
5. The trunk image analysis means that after the opponent wood is shot, a red calibration line automatically appears on a software interface, and the trunk is divided into a plurality of parts. Adjusting a red calibration line of the software along the vertical direction, and segmenting the trunk in the shot picture; and sliding in the horizontal direction, determining left and right intersection points of each trunk and the calibration line, and manually acquiring the coordinates of the intersection points of each section of the trunk and the calibration line.
6. The trunk data calculation means that the software automatically calculates the diameter of the bottom of each part of the trunk and the coordinates of the middle point of the calibration line according to the following formula, and calculates the vertical height between the two calibration lines according to the formula.
H=L(tanα+tanβ)
H is the tree height. (P)Tx,PTy) And (P)Bx,PBy) The coordinates of the top and bottom of the tree, respectively. DiIs the diameter of the base of the ith point of the trunk. h isiRepresenting the vertical height of two adjacent calibration lines on the trunk at point i. (P)iLx,PiLy) And (P)iRx,PiRy) And coordinates of intersection points of the two sides of the trunk at the ith and fixed marking lines. (P)iTx,PiTy) And (P)iBx,PiBy) Respectively represents the coordinates of the middle points of two adjacent calibration lines of the trunk at the ith point.
7. The calculation of the standing timber volume refers to approximating each part of the whole tree by using FSID according to the geometry and characteristics of the trunk. The whole trunk is divided into n sections, and the average section length is 0.5-3 m. The crown is considered to be a cone and the portion below the crown is considered to be the frustum of the cone. Finally, the volume of each segment is added to obtain the volume of the whole standing wood. The specific algorithm is shown in the following formula. The calculation formula of the conical frustum volume is as follows:
(i ═ 1,2, …, n) is the volume of the frustum of a cone. h is the height of the segmented frustum of the cone. DUIs the cross-sectional diameter of the top surface of the conical frustum DDIs the cross-sectional diameter of the base of the conical frustum. The cone volume calculation formula is:
Vni.e. the volume of the cone. h isn-1Is the height of the cone (i.e. the length of the crown), Dn-1The diameter of the base of the cone (i.e. the diameter of the lowermost part of the cone). The total volume formula V can be represented by the following formula:
compared with the prior art, the method has the following advantages:
1. the method is used for analyzing and obtaining the volume of the single standing tree, has no damage to the tree, is suitable for 5-30 m arbors and is suitable for small-range forest investigation.
2. The method is based on FSID and a double-point observation method, and the precision meets the forest investigation requirement.
3. The invention develops a portable forest intelligent tree measuring instrument, and the economic cost is lower.
4. The invention is convenient for popularization and easy and fast to operate.
Description of the drawings
The invention is further illustrated with reference to the figures and examples.
FIG. 1 is a schematic view of an intelligent tree tester for forest survey;
FIG. 2-software schematic diagram of equipment kit;
FIG. 3 is a schematic diagram of the principle of observation of standing trees by FSID;
FIG. 4-schematic view of a standing wood volume model;
FIG. 5-schematic view of a photography mode combining multiple photographs;
h is the tree height HiIs the height of the trunk section DiThe diameter of the trunk section, L is the horizontal distance from the observation point to the standing tree. Alpha is the angle between the observation site and the top of the tree, beta is the angle between the observation site and the bottom of the tree, i is 1,2, …, n, where Dn=0。
The fifth embodiment is as follows:
the method is based on a double-point observation method and FSID (offset frequency identification) to measure the volume of the single standing tree, and comprises the following specific implementation modes:
1. selecting a test standing tree, and erecting an intelligent tree tester for forest investigation. Forest survey software is turned on and the horizontal distance L between the instrument and the station is measured.
2. And taking pictures, and merging the pictures if the observation conditions are difficult. And measuring an included angle alpha between the observation station and the top of the tree, an included angle beta between the observation station and the bottom of the tree, and a plurality of intersection points of the calibration line and the trunk in the standing tree photo.
3. The method comprises the steps of carrying out standing tree trunk analysis through App, dividing a trunk into a plurality of sections, obtaining left and right intersection points of each trunk and a calibration line, and manually obtaining intersection point coordinates of each section of the trunk and the calibration line.
4. And carrying out tree trunk data processing by using the obtained measurement parameters, and finally carrying out standing timber volume calculation to obtain the single standing timber volume.
5. According to the method, on an Android Studio 3.3 development platform, forest survey software developed by using a Java programming language is used, and an Open Source Computer Vision Library (OpenCV) interface is called by using a Java Native Interface (JNI) technology to process the standing tree photo. The software can measure the angle, the distance and the volume of the standing tree in real time. The user can also carry out the concatenation of many photos to the standing tree of shooing according to the actual work needs. The user can adjust the instrument according to the actual measurement environment and the measurement work requirement, and angle data, distance data and pictures are processed through information transmission and integration among the modules. All measurement parameters, such as generated volume data, standing photographs, etc., can be stored or derived (. csv) in SQLite database format, or data can be derived in spreadsheet form using JXL technology.