Merge branch 'main' of gitlab.com:Trygve/contour-creator

README.md

@@ -5,7 +5,9 @@ This is our project for the INF205: Resource-efficient programming course
 - [x] Read .tif file into memory using gdal
 - [ ] Run the marching squares algorithm and produce a "casemap"
 - [ ] Use a lookuptable to produce a vector file from the "casemap"
-
+## Dependencies
+- GDAL
+- OpenMP
 ## How to build:
 ```
 mkdir build
@@ -13,4 +15,4 @@ cd build
 cmake -DCMAKE_BUILD_TYPE=Debug ..
 cmake --build . --parallel
 ```
-Then you can run `./contour-creator`
+Then you can run `./contour-creator PATH/TO/HEIGTHMAP.TIF`

documentation/report.md

@@ -1,14 +1,17 @@
 ## Group: Trygve and Esther
-# What problem will you be working on in your programming project? 
-We will be implementing a marching squares algorithm to produce a contour map from a heightmap image file.
-
+# Functionality
+Our program usees the marching squares algorithm to create a vector contour map from a raster heightmap.
+The cli interface is `gdal_contour [OPTIONS] <src_filename> <dst_filename>` with these options:
+```
+-i <elevation intervall>    Interval between contours
+-f <format>                 Fileformat to output
+```
+# data structure and input/output
+![](ER_diagram.svg)
 # Responsibilities:
 Esther will create the algorithm itself with multitreading. This will essentially be a function that takes a grid of pixels as input and returns a similar grid of cells.
 Trygve will take care of reading in the tiff file into our own datastructure and creating a vector image from the output of the algorithm.
 
 # How do you plan to make it easily verifiable that your objectives are reached?
 We can compare against the `gdal_contour` cli program which is a implementation widely used in other software. We can compare speed, memory usage and the result itself.
-Each step in our program also produces a output which we can be worked on and evaluated independently.
-
-# ER diagram:
-![](ER_diagram.svg)
+Each step in our program also produces a output which we can be worked on and evaluated independently.

src/HeightMap.cpp

@@ -1,25 +1,26 @@
+#include <stdexcept>
+
 #include <gdal/gdal.h>
 #include "gdal/gdal_priv.h"
-#include <iostream>
-
-#include <stdfloat>
+#include <gdal/gdal_frmts.h>
 
 #include "HeightMap.hh"
 
 HeightMap::HeightMap(const char* filepath) 
 {
-    // Open the file with some gdal nonsense:
+    // Open the file with gdal:
     const GDALAccess eAccess = GA_ReadOnly;
     GDALDatasetUniquePtr file;
-    GDALAllRegister();
+    GDALRegister_GTiff();
     file = GDALDatasetUniquePtr(GDALDataset::FromHandle(GDALOpen( filepath, eAccess )));
     if( !file )
     {
-        std::cout << "Could not open tiff file!"; // handle error
+        throw std::runtime_error("Could not open tif file!");
     }
     
     // The heigthmap only has one band
     auto band = file->GetBands()[0];
+    
     // write the attrributes
     this->width = band->GetXSize();
     this->height = band->GetYSize();
@@ -27,10 +28,10 @@ HeightMap::HeightMap(const char* filepath)
     this->max = band-> GetMaximum();
     
     this->data = (float *) CPLMalloc(sizeof(float)*width*height);
-    band->RasterIO( GF_Read, 0, 0, width, height,
+    CPLErr error = band->RasterIO( GF_Read, 0, 0, width, height,
                     this->data, width, height, GDT_Float32,
                     0, 0 );
-    
+    if (error) { throw std::runtime_error("Could not read tif file!"); }
     band->FlushCache();
 }
 float HeightMap::get_pixel(int x, int y)
@@ -38,4 +39,4 @@ float HeightMap::get_pixel(int x, int y)
     // all the pixels are in an array of floats from left to right, top to bottom
     int offset = ((this->width * y) + x);
     return *(this->data + offset);
-} 
+}