This patent application claims priority to Korean Application No. 10-2014-0076702, filed Jun. 23, 2014, the entire teachings and disclosure of which are incorporated herein by reference thereto.
1. Field of the Invention
The present invention relates to a sonar-image-simulation method for image prediction of an imaging sonar and an apparatus that uses the sonar-image-simulation method.
2. Description of the Related Art
Generally, underwater object recognition has great potential for automatic object detection, environment monitoring, and safety inspection and maintenance of underwater structure. Additionally, for example, it is applicable to underwater mine tracking via autonomous underwater vehicles. Although underwater optical sensors for this end provide the highest-resolution underwater images, their visibility is generally limited due to turbidity of the water. Compared to optical vision, imaging sonar, which uses acoustic signals to visualize underwater environments, has a long visual range even in turbid water and thus it is treated as an alternative approach for underwater object recognition.
However, a display mechanism of the imaging sonar is completely different from that of optical cameras, which is generally described by a pinhole model. Imaging sonar generates a sonar image by measuring a distance and reflective acoustic signal from a target object. That is, even an object with a well-known size and shape may appear quite different from an intuitively predicted shape in the sonar image. Moreover, the shape of an object in the image may vary completely even with a small change in the view point.
With considering the aforementioned points, if the sonar image of an object to be detected is known in advance, the detection may become reliable to perform. Therefore, a simulator for advance prediction of sonar images is needed.
Accordingly, the present invention has been made keeping in mind the above problems occurring in the prior art. The object of the present invention is to provide a sonar-image-simulation method for image prediction of an imaging sonar to predict a sonar image of a target object and an apparatus using the sonar-image-simulation method.
In order to accomplish the above object, the present invention provides a sonar-image-simulation method for image prediction of an imaging sonar, in which it sets a model of ultrasound beams of the imaging sonar, a model of a target object, and a model of the background on which the target object is placed as multiple straight lines, multiple polygons, and an infinite plane, respectively. These models are just elementary geometry that can be easily stated by mathematical equations. The sonar-image-simulation method simulates image data by calculating intersection points of the multiple straight lines representing the ultrasound beams with the multiple polygons representing the target object or the infinite plane, and deriving image data based on the calculated intersection points to construct a simulated sonar image.
The setting may include a first setting operation that sets a total number of the straight lines and their directions as the parameters of the ultrasound beam model and a position and an orientation of the imaging sonar and a second setting operation that sets multiple polygon meshes as the parameters of the target object model.
In the second setting operation, the polygons' shape may be a triangle.
The setting may include receiving parameters of the model of the target object in the form of a computer-aided design (CAD) file.
The simulating may include transforming coordinates of the multiple polygons from a global coordinate system to a local coordinate system, calculating the intersection points of the multiple straight lines and the coordinate-transformed multiple polygons or the infinite plane, and the image data depending on a position of the calculated intersection points in the local coordinate system.
The intersection points may include determining whether the intersection points are inside the polygon area being examined.
The image data may include calculating the intersection point's absolute distance r, and each straight line's an azimuth angle θ.
The calculating the image data may include selecting a straight line having the higher intensity between two straight lines if the two straight lines have the same distance and azimuth angle but different elevation angle.
The calculating the image data may include selecting a intersection point having a smallest distance among two or more intersection points if the two or more intersection points are calculated on one straight line.
The calculating the image data may include determining an intensity corresponding to the intersection points according to the type of an object which the straight line meets.
The calculating the image data may include determining the intensity to be highest if the straight lines meet the polygons, to be lowest if the straight lines meet neither the polygons nor the infinite plane, and to be intermediate if the straight lines meet the infinite plane.
The image-prediction simulation method may include displaying the simulated sonar image.
The present invention also provides an image-prediction simulation apparatus executing the image-prediction simulation method described above.
The above and other objects, features and advantages of the present invention will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings, in which:
Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings to allow those of ordinary skill in the art to easily carry out the embodiments. The present invention may be implemented in various forms, without being limited to the embodiments described herein. In the drawings, parts that are not relevant to a description of the present invention will not be provided for clarity, and the same reference numerals are used throughout the different drawings to designate the same or similar components.
Referring to
Parameter setting unit 110 sets parameters of the ultrasound beam model of the imaging sonar, parameters of the target object model, and those of the background model. Parameter setting unit 110 may include imaging sonar parameter setting unit 112 and object parameter setting unit 114.
More specifically, imaging sonar parameter setting unit 112 sets the total number of straight lines and their directions representing the ultrasound beams and a position and orientation of the imaging sonar. In the imaging sonar, the ultrasound beams are emitted in the form of a nearly-planar beam as illustrated in
As illustrated in
In
An equation of each straight line may be defined in the local coordinate system, as follows:
where p represents an arbitrary point on the straight line, and {right arrow over (v)} represents a normalized direction vector. The origin of the coordinate system is considered as a source of beams.
Referring back to
Each polygon can be defined as an equation of plane with boundary which is simple to manipulate. For example, a triangular polygon needs three points for itself to be defined. The plane of a polygon implies a part of the plane passing those three points, and its boundary is defined by the lines connecting the points. Therefore, instead of dealing with complex surface equation of objects, we can simplify the necessary computation by using polygons' plane equations.
As illustrated in
{right arrow over (N)}=({right arrow over (p2)}−{right arrow over (p1)})×({right arrow over (p3)}−{right arrow over (p1)})
The equation of the plane may satisfy the following condition:
{right arrow over (N)}·({right arrow over (p)}−{right arrow over (p1)})=0,
where the point p is an arbitrary point on the plane.
Parameter setting unit 110 may set the parameters of the target object described above, but may also receive the parameters of the target object from outside in the form of a computer-aided design (CAD) file. In this case, object parameter setting unit 114 may use the CAD file to set the parameters of the target object.
Referring back to
Simulator 120 simplify the reflection of the ultrasound beams from the surface of objects by computing an intersection point of a straight line with a polygon, and calculating the absolute distance of the intersection point from origin of the local coordinate system.
In
The imaging sonar is a sensor that can be attached to a ship or an underwater vehicle, so that the imaging sonar may have translation or rotation motion. Thus, to describe the ultrasound beams emitted from the imaging sonar, a local (imaging sonar) coordinate system is defined.
Simulator 120 transforms multiple polygon coordinates, more specifically, the coordinates of polygons' vertices which defined in the global coordinate system into those in the local coordinate system. After the unification of coordinate system simulator 120 can perform the calculation of the intersection points, more specifically, the coordinates of the intersection points of multiple polygons with straight lines.
Computation of the intersection points includes the decision whether a straight line passes any polygons. If the straight line meets any polygon at some point (the intersection point), simulator 120 find the point's coordinate. This computation can be considered as a reflection model of the ultrasound beam.
Calculation of the intersection point may use the following method for efficient calculation because there are a number of polygons and straight lines. For each polygon, the points consisting of the polygon can be described with the azimuth angle and elevation angle in the local coordinate system. Therefore, by setting those coordinates as boundaries, simulator 120 can find the azimuth angle range and elevation angle range that cover the polygon's positional domain. The ranges imply possible and potential angle ranges that some straight lines should belong to in order to meet the polygon. That is, only the straight lines within these ranges are considered to find the intersection points. In this way simulator 120 can select potential straight line candidates for each polygon, and as a result efficiency of computation is increased.
If the intersection point is calculated, a distance from the source, i.e., an origin of the local coordinate system to the intersection point may be calculated, and the distance value may be used to construct a sonar image in such a way the real imaging sonar does.
If one straight line meets two or more polygons and therefore obtain two or more intersection points, simulator 120 select the intersection point that is closest to the origin. In physical interpretation, this practice reflects the fact that ultrasound beam returns right after initial collision.
At this time, the important point is that intersection points should be inside the polygon boundary defined by the polygon vertices.
(({right arrow over (p′2)}−{right arrow over (p′1)})×({right arrow over (p)}−{right arrow over (p′1)}))·{right arrow over (N′)}≧0
(({right arrow over (p′3)}−{right arrow over (p′2)})×({right arrow over (p)}−{right arrow over (p′2)}))·{right arrow over (N′)}≧0
(({right arrow over (p′1)}−{right arrow over (p′3)})×({right arrow over (p)}−{right arrow over (p′3)}))·{right arrow over (N′)}≧0
where p1′, p2′, p3′ represent the positions of vertices of the polygon, p the intersection point, and N′ normal vector of the polygon. Once the intersection point is identified inside the polygon, simulator 120 accepts the point validated. Otherwise, simulator 120 ignores the intersection point.
After computing the intersection points for every polygons and straight lines, simulator 120 computes again the intersection points' absolute distance (or Euclidian distance) that is essential component to construct simulated sonar images. The simulator 120 calculates each straight lines' azimuth angle θ, and an elevation angle φ. In this case, the emulator 120 may select a straight line having the higher intensity between two straight lines if two straight lines have the same distance and azimuth angle, but different elevation angles.
Simulator 120 determines an intensity value corresponding to the calculated intersection point depending on type of objects the line meets. For example, simulator 120 may determine the intensity value to be highest if the straight line meets the object polygons; the intensity value to be lowest if the straight line does not meet either the polygon or the infinite plane (shadow case); and the intensity value to be intermediate if the straight line meets the infinite plane representing background. In this way, simulator 120 according to an embodiment of the present invention classifies the intensity values corresponding to the intersection point into three types.
Sonar image constructing unit 130 maps the image data obtained from simulator 120 into an image plane to construct a simulated sonar image.
Sonar image is constructed by referring each line's specific information such as its azimuth angle, the intersection point's absolute distance value, and corresponding intensity value. The sonar image plane consists of two axis indicating azimuth angle (X) and absolute distance (Y) (
Referring back to
With the above-described structure, the sonar-image-simulation apparatus 100 according to an embodiment of the present invention can simulate any object or underwater environments by constructing simulated sonar images when appropriate 3D models are given. Simulated sonar images have no noise that is inevitable in real sonar images. In addition, the computing efficiency is good enough to be used in a real time basis. These characteristics make prediction of the edge shape of a target object easy.
Hereinafter, a sonar-image-simulation method for image prediction of an imaging sonar will be described with reference to
Referring to
Next, the model of the target object is set to multiple polygons at step S902. As the parameters of the model of the target object, the number of polygons and their vertices' coordinate are defined in the global coordinated system. Normally, the parameters of target objects are read from 3D CAD files.
Next, coordinates defining the multiple polygons are transformed from the global coordinate system into the local coordinate system at step S903. It is because the coordinate systems defining the multiple polygons and straight lines are different. By this coordinate system unification, necessary mathematical manipulation can be done.
The intersection points of the straight lines with the multiple polygons or background are calculated at step S904.
At step S905, it is determined whether the obtained intersection point is inside the polygon or polygon boundary; if the intersection point is not inside the polygon, the intersection point is ignored, or if the intersection point is inside the polygon, image data is calculated from the obtained intersection points at step S906.
The calculated image data is mapped to the image plane at step S907 by referring each line's specific information such as its azimuth angle, the intersection point's absolute distance value, and corresponding intensity value. For each and every line, it finds the position corresponding to the line's azimuth angle and distance value in the image plane. In that pixel position, the intensity value is written. The shape of the image plane is transformed into a sector form at step S908. For example, as illustrated in
Next, the constructed sonar image is displayed at step S908. The sonar image may be displayed through, but not limited to, a general display device connected to a computer.
The foregoing methods may be implemented by the sonar-image-simulation apparatus 100 as illustrated in
Without these differences, the overall shape and size of the target object is very similar in the real sonar image and the simulated images.
As described above, the sonar-image-simulation method and apparatus for image prediction of an imaging sonar can simulate any object or underwater environments by constructing simulated sonar images when appropriate 3D models are given. Simulated sonar images have no noise that is inevitable in real sonar images. In addition, the computing efficiency is good enough to be used in a real-time basis. These characteristics make prediction of the edge shape of a target object easy.
Although the preferred embodiments of the present invention have been disclosed for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the invention as disclosed in the accompanying claims.
Number | Date | Country | Kind |
---|---|---|---|
10-2014-0076702 | Jun 2014 | KR | national |