I have a java applet that allows users to import a jpeg and world file from the local system. The user can then "click" draw lines on the image that was imported. Each endpoint of each line contains a set of X/Y and Lat/Long values. The XY is standard java coordinate space, the applet uses an affine transform calculation with the world file to determine the lat/long for every point on the canvas.
I have a requirement that allows a user to type a distance into a text field and use the arrow key to draw a line in a certain direction (Up, Down, Left, Right) from a single selected point on the screen. I know how to determine the lat/long of a point given a source lat/long, distance, and bearing.
So a user types "100" in the text field and presses the Right arrow key a line should be drawn 100 feet to the right from the currently selected point.
My issue is I don't know how to convert the distance( which is in feet ) into the distance in pixels. This would then tell my where to plot the point.
tcarobruce,
You are correct. The inverse transform algorithm is what I needed. Since I use java I was able to replace my "home made" transform algorithm with the java.awt.AffineTransform object which has an inverse transform function.
This seems to have solved my issue.
Thanks.
I guess you are certain your users are always uploading a raster image that is in the lat/lon wgs84 projection? Because in that case you can set a fixed coordinate transformation.
If you consider ever digitizing images from other sources with other projections, you might want to take a look at the open source geotools library: http://www.geotools.org/
Related
I aim to make an object detection model and I labelled data with a square box
If I label the images with polygon, will it be better than square?
(labelling on image of people wearing safety helmet or not)
I did try label with polygon shape on a few images and after export txt file for YOLO
why it has only 4 points in the text file as same as labelled with a square shape
how those points will represent an area that I label accurately?
1 0.573748 0.018953 0.045332 0.036101
1 0.944520 0.098375 0.108931 0.167870
You have labeled your object in a polygonial format, but when you had made a conversion to YOLO-format the information in the labelings has reduced. The picture below shows how I suppose has happend;
...where you have done polygon shape annotation (black shape). But, the conversion has "searched" the smallest x-value from the polygonial coordinate points and smallest y-value from corresponding polygonial coordinate points. And, those are the "first two" values of your YOLO-format. The same logic has happend with the "width" and "heigth" -parameters.
A good description about the idea behind the labelling and dataset is shown in https://www.youtube.com/watch?v=h6s61a_pqfM.
In short; for your purpose (for efficiency) I propose you make fast & convenient annotation using rectangles only - no time consuming polygon annotation.
The YOLO you are using very likely only has square annotation support.
See this video showing square vs polygon quality of results for detection, and the problem of annotation time required to create custom data sets.
To use polygonal masks can I suggest switching to use YOLOv3-Polygon or YOLOv5-Polygon
So the software I am using accepts 3D objects in the form of contours or .stl files. The contours I have are along the z-plane(each plane has a unique z). I have had to modify the contours for my experiment and now the contours do not have a unique z for each plane(they are now slightly angled wrt z=0 plane).
The points represent the edges of the 3D object. What would be the best way to take this collection of points and create a .stl file?
I am relatively new to working with python and 3D objects, so any help, pointers or suggestions would be much appreciated.
Edit: I have the simplices and verticies using the Delaunay(), but how do I proceed next?
The co-ordinates of all points are in this text file in the format "x y z".
So after seeking an answer for months and trying to use Meshlab and Blender I finally stumbled across the answer using numpy-stl. Hopeful that it will help others in a similar situation.
Here is the code to generate the .STL file:
from stl import mesh
num_triangles=len(fin_list)
data = np.zeros(num_triangles, dtype=mesh.Mesh.dtype)
for i in range(num_triangles):
#I did not know how to use numpy-arrays in this case. This was the major roadblock
# assign vertex co-ordinates to variables to write into mesh
data["vectors"][i] = np.array([[v1x, v1y, v1z],[v2x, v2y, v2z],[v3x, v3y, v3z]])
m=mesh.Mesh(data)
m.save('filename.stl')
The three vertices that form a triangle in the mesh go in as a vector that define the surface normal. I just collected three such vertices that form a triangle and wrote them into the mesh. Since I had a regular array of points, it was easy to collect the triangles:
for i in range(len(point_list)-1):
plane_a=[]
plane_b=[]
for j in range(len(point_list[i])-1):
tri_a=[]
tri_b=[]
#series a triangles
tri_a.append(point_list[i+1][j])
tri_a.append(point_list[i][j+1])
tri_a.append(point_list[i][j])
#series b triangles
tri_b.append(point_list[i+1][j])
tri_b.append(point_list[i+1][j+1])
tri_b.append(point_list[i][j+1])
#load to plane
plane_a.append(tri_a)
plane_b.append(tri_b)
group_a.append(plane_a)
group_b.append(plane_b)
The rules for choosing triangles for creating a mesh are as follows:
The vertices must be arranged in a counter-clock direction.
Each triangle must share two vertices with adjacent triangles.
The direction normal must point out of the surface.
There were two more rules that I did not follow but it still worked in my case:
1. All coordinates must be positive(In 1st Quadrant only)
2. All triangles must be arranged in an increasing z-order.
Note: There can be two kinds of .STL file formats: Binary and ASCII. numpy-stl writes out in the binary format. More info on STL files can be found here.
Hope this helps!
I am trying to extract text from an image, but within a certain area of the image and not the entire image.
I have already been able to detect where the objects of interest are and get their coordinates. Though I do not know where to start when extracting text from a specific area.
I'm using the code from this example:
https://www.codingame.com/playgrounds/38470/how-to-detect-circles-in-images
It is able to detect the circles, but I want to take it one step further and extract the numbers from the circles and tag them to their corresponding coordinate.
I'm using this example to learn how to do something similar myself, but I'm really more interested in deciding the search in a set area.
Most image processing libraries support the concept of ROIs (region of interest) or AOIs (area of interest).
The idea is to restrict processing to a subset of pixels that are usually selected by defining geometric shapes like rectangles, polygons, circles within the image coordinate system.
You can fix this issue by first cropping the image using your coordinates and try to extract text from it.
I have a program which is able to parse and interprets OBJ file format in an OpenGL context.
I created a little project in Blender containing a simple sphere with 'Hair' particles on it.
After conversion (separating particules from the sphere) my particles form a new mesh. So I have two meshes in my project (named 'Sphere' and 'Hair'). When I want to export the mesh 'Sphere' in an OBJ file (File/export/Wavefront (.obj)), selecting 'include Normals', after exportation, the file contains all informations about normals (ex: vn 0.5889 0.14501 0.45455, ...).
When I try to do the same thing with particles, selecting 'include Normals' too, I don't have normals in the OBJ file. (Before the exporting I have selected the right mesh.)
So, I don't unsterstand why normals properties are not exported for mesh of type particles.
Here's above the general Blender render of my hair particules. As you can see all particules have a reaction with the light. So Blender use normals properties for thoses particules.
And now, the picture above shows (in Blender 'Edit mode' -> after conversion) that particules are formed of several lines. In my opengl program I use GL_LINES to render the same particules. I just want to have normals information to manage light properties on my particules.
Do you have an idea how to export normals properties for particules meshes?
Thanks in advance for your help.
You are trying to give normals to lines. Let's think about what that means.
When we talk about normal vectors on a surface, we mean "pointing out of the surface"
For triangles, when we define one side to be the "front" face, there is exactly one normal. For lines, any vector perpendicular to the line counts as a normal - there are infinite and any one will "do".
What are some reasons we care about normals in graphics?
Lighting: e.g. diffuse lighting is approximated by using the dot product of the normal with the incident light vector. This doesn't apply to hair though!
Getting a transformation matrix: for this you can pick any normal (do you want to transform into hair-space?)
In short: you either can pick any perpendicular vector for your normal (it's easy to calculate this) or just not use normals at all for your hair. It depends on what you are trying to do.
I got my country lat/long boundaries from koordinates.com. Now I want to fill in the interior with dots.
Since the file I have is KML, I was thinking of converting the coordinates to cartesian using the NetTopologySuite.
I do not want a polygon overlay. I want to generate dots/coordinates for the polygons interior - ideally at a density of my choosing.
I have seen algorithms like this one, http://alienryderflex.com/polygon_fill/. Is there a library that will do this for me? Alternatively, can someone share code?
Ultimately, I will convert the dot coordinates back to lat/long and populate a globe like this one
http://code.google.com/p/webgl-globe/
I'm affraid GIS isn't my area of expertise, but I've got two ideas:
Generate a set of random points. You can use a Point-In-Polygon function to determine if you're points are in the right place.
You can use a rectangle grid of points and use a 'resolution' to determine how many points there will be and how close. You can offset the grid positions to make them look more random if you need to. You'll check if the point inside the bounding rectangle of your polygon is inside the polygon or not.
Notice that the webgl-globe example uses a grid of points(similar to point(2)) converted to spherical coordinates.
Both ideas is kind of similar, only the points distribution differs.
You can find a roughly related implementation I did using actionscript here,
but I would also suggest asking on the GIS site.