CN110689563A

CN110689563A - Data processing method for extracting illegal building information in remote sensing image

Info

Publication number: CN110689563A
Application number: CN201910925268.9A
Authority: CN
Inventors: 王雪; 张珊珊; 欧伟健
Original assignee: Foshan University
Current assignee: Foshan University
Priority date: 2019-09-27
Filing date: 2019-09-27
Publication date: 2020-01-14

Abstract

The invention discloses a data processing method for extracting illegal building information in a remote sensing image, and relates to the technical field of unmanned aerial vehicle remote sensing monitoring. The method comprises data collection and preparation, wherein digital orthographic image achievements which are manufactured and completed in N years ago in an area to be researched are obtained, and preparation work of an unmanned aerial vehicle aerial photography system is performed in advance to perform real-time aerial photography work, and the unmanned aerial vehicle aerial photography: and carrying out aerial photography work on the area to be researched according to the determined monitoring range and the determined flight plan, and acquiring preliminary high-resolution remote sensing ortho-image data after the aerial photography is finished. According to the method, an unmanned aerial vehicle remote sensing technology is utilized to obtain an orthoscopic image of a region to be researched, a series of processing is carried out on the image, so that a building remote sensing image with high spatial resolution is finally obtained, illegal construction is found out by comparing remote sensing images in the past year, and a set of efficient and accurate illegal building monitoring technical scheme is formed.

Description

Data processing method for extracting illegal building information in remote sensing image

Technical Field

The invention belongs to the technical field of unmanned aerial vehicle remote sensing monitoring, and particularly relates to a data processing method for extracting illegal building information in a remote sensing image.

Background

In recent years, the remote sensing technology of the unmanned aerial vehicle is rapidly developed and widely applied. For example, the unmanned aerial vehicle technology is widely applied to land utilization planning, land reclamation, rural land contract registration, fine detection of mine geological disasters, land consolidation construction monitoring and investigation and evaluation of mountain torrent disasters. .

The conventional traditional means for monitoring illegal buildings is still carried out through manual investigation and field visit, or data are acquired through satellite remote sensing images and common aviation, but the following defects are high cost, long period for acquiring data and lack of maneuverability and flexibility.

Disclosure of Invention

The invention aims to provide a data processing method for extracting illegal building information in a remote sensing image, which solves the problems of high cost and poor precision of the existing data processing method for illegal building information by monitoring illegal buildings by adopting an unmanned aerial vehicle and optimizing the extraction method for the illegal buildings.

In order to solve the technical problems, the invention is realized by the following technical scheme:

the invention relates to a data processing method for extracting illegal building information in a remote sensing image, which comprises the following steps:

SS001, data collection and preparation: acquiring a finished digital orthophoto result of an area to be researched N years ago, and performing preparation work of an unmanned aerial vehicle aerial photography system in advance so as to perform real-time aerial photography work;

SS002, unmanned aerial vehicle aerial photography: carrying out aerial photography work on a region to be researched according to a determined monitoring range, a determined flight plan, a determined image control point and a determined course overlapping degree, acquiring initial high-resolution remote sensing ortho-image data after the aerial photography is finished, timely checking the image data after the initial high-resolution image data is acquired, carrying out supplementary photography and retaking of related regions according to imaging conditions, and finally acquiring a complete high-resolution remote sensing ortho-image of the region to be researched;

SS003, image preprocessing: importing the coordinates of the image control points in the SS002 step into Pix4D Mapper unmanned data processing software, positioning the image control points in the image in a manual interaction mode, selecting a standard plane coordinate system in the Pix4D Mapper software, selecting a corresponding projection band to perform Gaussian projection 3-degree banding of the image data, automatically identifying the characteristic points by using an automatic empty three-dimensional processing module in the Pix4D Mapper, performing encryption calculation on the characteristic points and the control points, acquiring high-precision three-dimensional coordinates, and outputting a precision quality report of the empty three-dimensional processing, thereby finally acquiring high-precision exterior orientation elements and initial matching point clouds of the image;

SS004, point cloud encryption: combining the high course overlapping degree of the unmanned aerial vehicle image in the SS002, obtaining high-precision exterior orientation elements of the image and the initial matching point cloud obtained in the SS003 step through a high-precision dense matching algorithm to obtain a large number of pixel-level matching points, and generating a landform dense point cloud, thereby finally producing a digital surface model;

SS005, digital surface model and orthoimage production: the method comprises the steps that a DSM and DOM production module of Pix4D Mapper software is used for manufacturing a digital surface model of a remote sensing ortho image, after the digital surface model is generated, the digital ortho image is generated by a digital differential correction method, automatic ortho correction is carried out on the image through Pix4 DMaper software after the digital ortho image is manufactured, the optimal mosaic line is automatically searched for and spliced into an ortho image, meanwhile, the mosaic line is manually adjusted in an area with an unsatisfactory splicing effect, the mosaic line is located in an area with small projection deformation such as a road, and the ortho image effect is improved;

SS006, image color homogenizing treatment and splicing: adjusting the color, brightness and contrast of the image and homogenizing the image, and simultaneously dividing the region to be researched into blocks so as to make an orthographic image;

SS007, coordinate transformation and framing numbering of images: after generating an orthoimage result, converting the orthoimage result of the research area through a coordinate conversion system, framing and labeling the orthoimage in a specified form after coordinate conversion, framing and numbering the image result of each block according to a 1:2000 scale through orthoimage framing software, and simultaneously automatically generating a picture and frame combination table;

SS008, extracting an illegal map spot: performing pattern spot extraction on the manufactured digital orthographic image result in an ArcGIS software environment, reading and displaying an unmanned aerial vehicle image by utilizing the ArcGIS, drawing a picture-shaped element according to the boundary of a target area, creating a grid through a Fishnet tool, drawing a pattern spot by the grid, avoiding omission, and drawing a planar pattern spot along the boundary of a building under construction during drawing;

SS009, creation of a plaque attribute table: uniformly numbering the extracted suspected illegal drawing spots, calculating the area of the drawing spots, removing the drawing spots with the too small area, superposing a 1:500 topographic map, extracting the position information of the drawing spots from the drawing, and recording the position information into an attribute table;

SS010, making a suspected law enforcement guide drawing for illegal construction: in an ArcGIS software platform, overlapping suspected illegal construction pattern spots, a large-scale topographic map and unmanned aerial vehicle orthographic image data in a layering manner, uniformly typesetting and arranging, outputting a law enforcement guide map for each pattern spot, and naming and arranging according to a uniform rule;

SS011, establishing a pattern spot database: inputting attribute information of suspected illegal building pattern spots, such as pattern spot numbers, areas, addresses and the like into an ARCGIS database, then exporting and analyzing, and finally exporting data maps such as suspected illegal building area ratio maps, suspected illegal building area increase conditions and the like;

further, the optical image precision of the high-resolution remote sensing ortho image data in the SS002 step is 3-10 CM.

The invention has the following beneficial effects:

according to the method, the remote sensing technology of the unmanned aerial vehicle is utilized to obtain the ortho-image of the area to be researched, and a series of processing is carried out on the image, so that the building remote sensing image with high spatial resolution is finally obtained, illegal construction is found out by comparing the remote sensing images in the past year, and a set of efficient and accurate illegal building monitoring technical scheme is formed.

Of course, it is not necessary for any product in which the invention is practiced to achieve all of the above-described advantages at the same time.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used for describing the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without any creative effort.

FIG. 1 is a flow chart of a data processing method for extracting illegal building information in remote sensing images;

FIG. 2 is a schematic view of a ground monitoring screen for unmanned aerial vehicle operation;

FIG. 3 is a schematic diagram of the arrangement positions of image control points;

FIG. 4 is a schematic diagram of an initial matching point cloud;

FIG. 5 is a schematic diagram of a dense matching point cloud;

FIG. 6 is a diagram of a digital surface model;

FIG. 7 is a schematic diagram of an artificial adjustment of an orthoimage mosaic display;

FIG. 8 is a partial orthophotomap of a region to be studied;

FIG. 9 is an exemplary diagram of suspected violation patterns;

FIG. 10 is a suspected law enforcement guidance diagram for illegal construction;

FIG. 11 is a digital surface model view;

Detailed Description

The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1-11, the present invention is a data processing method for extracting illegal building information in a remote sensing image, including the following steps:

SS002, unmanned aerial vehicle aerial photography: according to a determined monitoring range, further implementing and refining an aerial photography specific region of the unmanned aerial vehicle, carrying out aerial photography work of the region to be researched in steps of a determined flight plan, a determined image control point and a determined course overlapping degree, after the aerial photography is finished, downloading an original aerial photography image and an aerial photography record file of the unmanned aerial vehicle of the region to be researched on site, checking the integrity of the data, namely obtaining preliminary high-resolution remote sensing orthographic image data, timely checking the image data after the preliminary high-resolution image data is obtained to check whether phenomena of blurring, missing photography, unclear image and the like exist, further carrying out supplementary photography and retaking of the related region as appropriate, carrying out timely supplementary photography on absolute leaks, relative leaks and other serious defects which appear in the aerial photography process, and carrying out timely supplementary photography on relative leaks and local defects of image internal industry encryption selection points such as clouds, and the like, Selecting a cloud shadow, a spot mark and the like, performing complementary shooting at the position of the leak, setting the length of a complementary shooting route to exceed one base line outside the leak, and finally obtaining a complete high-resolution remote sensing orthographic image of the area to be researched, wherein during shooting, an UX5 unmanned plane and an eBee unmanned plane are adopted for aerial shooting;

SS003, image preprocessing: importing the coordinates of the image control points in the SS002 step into Pix4D Mapper unmanned data processing software, positioning the image control points in an image in a manual interaction mode, selecting a standard plane coordinate system in the Pix4D Mapper software, selecting a 114-degree projection zone to perform Gaussian projection 3-degree banding of image data, automatically identifying the characteristic points by using an automatic empty three-dimensional processing module in the Pix4D Mapper, performing encryption calculation on the characteristic points and the control points to obtain high-precision three-dimensional coordinates, and outputting an empty three-dimensional processing precision quality report, wherein the precision quality report sequentially displays the contents of adjustment precision, camera calibration parameters, the number and distribution conditions of two-dimensional characteristic points and three-dimensional encryption points of the region, and finally obtains high-precision exterior position elements and initial matching of the image;

SS005, digital surface model and orthoimage production: the method comprises the following steps of utilizing a DSM (digital document model) and DOM (document object model) production module of Pix4D Mapper software to manufacture a digital surface model of a remote sensing ortho-image, after the digital surface model is generated, generating the digital ortho-image by a digital differential correction method, automatically carrying out ortho-correction on the image through the Pix4D Mapper software after the digital ortho-image is manufactured, automatically searching an optimal mosaic line, splicing the mosaic line into an ortho-image, manually adjusting the mosaic line in an area with an unsatisfactory splicing effect to avoid a high-rise building, enabling the mosaic line to be located in an area with small projection deformation such as a road and the like, and improving the ortho-image effect, wherein the step is very important for the framing of the image, so the following principles are followed when the mosaic line is edited:

the method is characterized in that linear ground objects such as ridges, roadside lines, water line and the like are arranged as much as possible, so that even if obvious boundary lines exist, image processing software is adopted to easily process the boundary lines at the later stage;

when the inlaid line is selected, the inlaid line needs to go around mountains and houses as far as possible and go along a road, a high-rise building needs to be avoided, the shielding of the high-rise building on other ground objects is reduced, more ground information is stored, and otherwise, the house is inverted or extruded mutually when DOM splicing is carried out between image pairs or routes;

important ground objects such as an electric tower are avoided as much as possible to ensure the integrity of the important ground objects;

the embedding line is as straight as possible, so that small-angle folding lines are avoided being selected, and the effect of taking a smooth curve which is approximate to a straight line is optimal;

the flaky water is bypassed as much as possible, and the color is kept uniform;

when the mosaic lines are laid in mountainous areas, places with obvious color boundaries are avoided as much as possible, and the colors are stored to be uniform and unobtrusive;

in the place where houses are concentrated, the houses can be bypassed as much as possible, and the houses can be walked along streets as much as possible without being bypassed, so that the integrity of the houses is preserved;

SS006, image color homogenizing treatment and splicing: the method comprises the steps of adjusting the color, the brightness and the contrast of an image and homogenizing the color, simultaneously dividing a region to be researched into blocks to manufacture an orthographic image, wherein the homogenizing treatment is to reduce the color tone difference among the images, ensure uniform color tone, moderate contrast and clear level, keep the color of a ground object undistorted, avoid the occurrence of a color homogenizing treatment trace, timely search the reasons for problems and phenomena such as dirty points, fuzziness, dislocation, distortion, deformation, garland, dirty points, scratches and the like of the image, process and retrospective optimization, and for the region with a large range, the method is influenced by the performance of unmanned aerial vehicle data processing software, cannot perform integral treatment at one time, perform orthographic image manufacturing on the blocks, ensure certain overlapping between adjacent blocks when the orthographic image is manufactured on the blocks, thus facilitating image splicing, like generally adopting a mosaic tool in ArcGIS or cGIS DAS, and for the blocks with complex boundaries, processing by adopting an embedding module of PhotoMod software, editing by an embedding line, and outputting an orthographic image of the whole area again;

SS007, coordinate transformation and framing numbering of images: after the orthographic image result is generated, because aerial photography processing software cannot directly output the coordinate result of the area to be researched, the orthographic image result of the area to be researched needs to be converted through a coordinate conversion system, the orthographic image is subjected to framing and labeling in a specified form after the coordinate conversion, the orthographic image result of each block is subjected to framing and numbering according to a 1:2000 scale through orthographic image framing software, and meanwhile, a picture combination table is automatically generated;

SS008, extracting an illegal map spot: the method comprises the following steps of extracting pattern spots of a manufactured digital ortho-image result in an ArcGIS software environment, reading and displaying an unmanned aerial vehicle image by utilizing the ArcGIS, drawing a picture-shaped element according to a target area boundary, creating a grid through a Fishnet tool, drawing pattern spots by the grid, avoiding omission, drawing a planar pattern spot along the building boundary under construction during drawing, and extracting the suspected illegal construction pattern spots according to the rule that the manufactured digital ortho-image result is collected N years ago in the SS001 step:

the image before N years has no buildings, and the image at the current period has buildings;

the image building before N years has no under-construction characteristics, and the image at the current stage has under-construction characteristics;

the image building has the construction characteristics before N years but is not extracted, and the image still has the construction characteristics at the present stage;

under the condition that no unmanned aerial vehicle image map reference exists in a partial region N years ago, the specific rule of extracting the image spots is as follows:

taking a building under construction as an extraction target;

the minimum spot area is 50 square meters;

buildings with obvious under-construction marks, such as scaffolds around the buildings, uncapped buildings and the like are judged as suspected map spots;

large building construction plates are also used for extracting suspected constructed pattern spots, and other requirements are flexibly adjusted according to requirements;

SS010, making a suspected law enforcement guide drawing for illegal construction: in an ArcGIS software platform, overlapping suspected illegal construction pattern spots, a large-scale topographic map and unmanned aerial vehicle orthographic image data in a layering manner, uniformly typesetting and arranging, outputting a law enforcement guide map for each pattern spot, naming and arranging according to a uniform rule, wherein except that a few dense pattern spots are merged into a map, other pattern spots correspond to the law enforcement guide map one by one;

wherein, the optical image precision of the high-resolution remote sensing ortho image data in the SS002 step is 3-10 CM.

In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A data processing method for extracting illegal building information in a remote sensing image is characterized by comprising the following steps:

SS001, data collection and preparation: acquiring a digital orthophoto result which is manufactured in an area to be researched by N years ago, and carrying out preparation work of an unmanned aerial vehicle aerial photography system in advance so as to carry out real-time aerial photography work;

SS002, unmanned aerial vehicle aerial photography: carrying out aerial photography work on a region to be researched according to a determined monitoring range, a determined flight plan, a determined image control point and a determined course overlapping degree, acquiring preliminary high-resolution remote sensing ortho-image data after the aerial photography is finished, timely checking the image data after the preliminary high-resolution image data is acquired, carrying out supplementary photography and retaking of related regions according to imaging conditions, and finally acquiring a complete high-resolution remote sensing ortho-image of the region to be researched;

SS003, image preprocessing: importing the coordinates of the image control points in the SS002 step into Pix4D Mapper unmanned aerial vehicle data processing software, positioning the image control points in the image in a manual interaction mode, selecting a standard plane coordinate system in the Pix4D Mapper software, selecting corresponding projection zones to conduct Gaussian projection 3-degree banding of image data, using an automatic empty three-dimensional processing module in the Pix4D Mapper to automatically identify the feature points, carrying out encryption calculation on the feature points and the control points, obtaining high-precision three-dimensional coordinates, and outputting a precision quality report of empty three-dimensional processing, so that high-precision exterior orientation elements and initial matching point clouds of the image are finally obtained;

SS004, point cloud encryption: combining the high course overlapping degree of the unmanned aerial vehicle image in the SS002, obtaining a large number of pixel-level matching points by a high-precision dense matching algorithm and combining the high-precision exterior orientation elements of the image and the initial matching point cloud obtained in the SS003 step, and generating a landform dense point cloud, thereby finally producing a digital surface model;

SS005, digital surface model and orthoimage production: the method comprises the steps that a DSM and DOM production module of Pix4D Mapper software is used for manufacturing a digital surface model of a remote sensing ortho image, after the digital surface model is generated, the digital ortho image is generated by a digital differential correction method, after the digital ortho image is manufactured, automatic ortho correction is carried out on the image through the Pix4D Mapper software, the optimal mosaic line is automatically searched for and spliced into an ortho image, meanwhile, for the area with the unsatisfactory splicing effect, the mosaic line is manually adjusted, so that the mosaic line is located in the area with small projection deformation, such as a road, and the ortho image effect is improved;

SS007, coordinate transformation and framing numbering of images: after generating an orthoimage result, converting the orthoimage result of the research area through a coordinate conversion system, framing and marking the orthoimage in a specified form after coordinate conversion, framing and numbering the image result of each block according to a 1:2000 scale through orthoimage framing software, and simultaneously automatically generating a picture and frame combination table;

SS008, extracting an illegal map spot: performing pattern spot extraction on the finished digital orthographic image in an ArcGIS software environment, reading and displaying an unmanned aerial vehicle image by utilizing the ArcGIS, drawing a planar element according to the boundary of a target area, creating a grid through a Fishnet tool, drawing pattern spots by the grid, avoiding omission, and drawing the planar pattern spots along the boundary of a building under construction during drawing;

SS009, creation of a plaque attribute table: uniformly numbering the extracted suspected illegal drawing spots, calculating the area of the drawing spots, removing the drawing spots with the too small area, superposing a 1:500 topographic map, extracting the position information of the drawing spots from the drawing, and recording an attribute table;

SS010, making a suspected law enforcement guide drawing for illegal construction: in an ArcGIS software platform, overlapping suspected illegal construction pattern spots, a large-scale topographic map and unmanned aerial vehicle orthographic image data in a layered mode, uniformly typesetting and laying out, outputting a law enforcement guide map for each pattern spot, and naming and arranging according to uniform rules;

2. the data processing method for extracting illegal building information from remote sensing images as claimed in claim 1, wherein the optical image precision of the high-resolution remote sensing ortho image data in the SS002 step is 3-10 CM.