codebase / stimlib

Commit 59071526e1543936be7509ca1d609b986ce0458a

Authored by David Mayerich
2 parents ee52e24f d9a27f16

Merge branch 'master' of git.stim.ee.uh.edu:codebase/stimlib

Showing 2 changed files with 26 additions and 48 deletions   Show diff stats
Inline Side-by-side
stim/iVote/ivote2.cuh
  Show/Hide comments   View file @5907152
... ... @@ -94,8 +94,7 @@ namespace stim {
94 94  
95 95 //this function performs the 2D iterative voting algorithm on the image stored in the gpu
96 96 template<typename T>
97   - void gpu_ivote2(T* gpuI, unsigned int rmax, size_t x, size_t y, bool invert = false, T t = 0, std::string outname_img = "out.bmp", std::string outname_txt = "out.txt",
98   - int iter = 8, T phi = 15.0f * (float)stim::PI / 180, int conn = 8, bool debug = false) {
  97 + void gpu_ivote2(T* gpuI, unsigned int rmax, size_t x, size_t y, bool invert = false, T t = 0, int iter = 8, T phi = 15.0f * (float)stim::PI / 180, int conn = 8, bool debug = false) {
99 98  
100 99 size_t pixels = x * y; //compute the size of input image
101 100 //
... ... @@ -129,43 +128,36 @@ namespace stim {
129 128 }
130 129 stim::cuda::gpu_local_max<float>(gpuI, gpuVote, conn, x, y); //calculate the local maxima
131 130  
132   - T* pts = (T*)malloc(bytes); //allocate memory on the cpu to store the output of iterative voting
133   - HANDLE_ERROR(cudaMemcpy(pts, gpuI, bytes, cudaMemcpyDeviceToHost)); //copy the output from gpu to the cpu memory
134   -
135   - T threshold;
136   - if (t == 0) threshold = stim::th_otsu<T>(pts, pixels); //if threshold value is not set call the function to compute the threshold
137   - else threshold = t;
138   -
139   - std::ofstream output; //save the thresholded detected seeds in a text file
140   - output.open(outname_txt);
141   - output << "X" << " " << "Y" << " " << "threshold" << "\n";
142   - size_t ind;
143   - for (size_t ix = 0; ix < x; ix++) {
144   - for (size_t iy = 0; iy < y; iy++) {
145   - ind = iy * x + ix;
146   - if (pts[ind] > threshold) {
147   - output << ix << " " << iy << " " << pts[ind] << "\n";
148   - pts[ind] = 1;
  131 + if (t > 0) {
  132 + T* pts = (T*)malloc(bytes); //allocate memory on the cpu to store the output of iterative voting
  133 + HANDLE_ERROR(cudaMemcpy(pts, gpuI, bytes, cudaMemcpyDeviceToHost)); //copy the output from gpu to the cpu memory
  134 +
  135 + T threshold;
  136 + threshold = t;
  137 +
  138 + size_t ind;
  139 + for (size_t ix = 0; ix < x; ix++) {
  140 + for (size_t iy = 0; iy < y; iy++) {
  141 + ind = iy * x + ix;
  142 + if (pts[ind] > threshold) {
  143 + pts[ind] = 1;
  144 + }
  145 + else pts[ind] = 0;
149 146 }
150   - else pts[ind] = 0;
151 147 }
  148 + HANDLE_ERROR(cudaMemcpy(gpuI, pts, bytes, cudaMemcpyHostToDevice)); //copy the points to the gpu
152 149 }
153   - output.close();
154   -
155   - HANDLE_ERROR(cudaMemcpy(gpuI, pts, bytes, cudaMemcpyHostToDevice)); //copy the points to the gpu
156   - stim::cpu2image(pts, outname_img, x, y); //output the image
157   -
  150 +
158 151 }
159 152  
160 153  
161 154 template<typename T>
162   - void cpu_ivote2(T* cpuI, unsigned int rmax, size_t x, size_t y, bool invert = false, T t = 0, std::string outname_img = "out.bmp", std::string outname_txt = "out.txt",
163   - int iter = 8, T phi = 15.0f * (float)stim::PI / 180, int conn = 8, bool debug = false) {
  155 + void cpu_ivote2(T* cpuI, unsigned int rmax, size_t x, size_t y, bool invert = false, T t = 0, int iter = 8, T phi = 15.0f * (float)stim::PI / 180, int conn = 8, bool debug = false) {
164 156 size_t bytes = x*y * sizeof(T);
165 157 T* gpuI; //allocate space on the gpu to save the input image
166 158 HANDLE_ERROR(cudaMalloc(&gpuI, bytes));
167 159 HANDLE_ERROR(cudaMemcpy(gpuI, cpuI, bytes, cudaMemcpyHostToDevice)); //copy the image to the gpu
168   - stim::gpu_ivote2<T>(gpuI, rmax, x, y, invert, t, outname_img, outname_txt, iter, phi, conn, debug); //call the gpu version of the ivote
  160 + stim::gpu_ivote2<T>(gpuI, rmax, x, y, invert, t, iter, phi, conn, debug); //call the gpu version of the ivote
169 161 HANDLE_ERROR(cudaMemcpy(cpuI, gpuI, bytes, cudaMemcpyDeviceToHost)); //copy the output to the cpu
170 162 }
171 163 }
... ...
stim/iVote/ivote2/local_max.cuh
  Show/Hide comments   View file @5907152
... ... @@ -10,30 +10,22 @@ namespace stim{
10 10  
11 11 // this kernel calculates the local maximum for finding the cell centers
12 12 template<typename T>
13   - __global__ void cuda_local_max(T* gpuCenters, T* gpuVote, int conn, size_t x, size_t y){
  13 + __global__ void cuda_local_max(T* gpuCenters, T* gpuVote, int conn, int x, int y){
14 14  
15 15 // calculate the 2D coordinates for this current thread.
16   - size_t xi = blockIdx.x * blockDim.x + threadIdx.x;
17   - size_t yi = blockIdx.y * blockDim.y + threadIdx.y;
  16 + int xi = blockIdx.x * blockDim.x + threadIdx.x;
  17 + int yi = blockIdx.y * blockDim.y + threadIdx.y;
18 18  
19 19 if(xi >= x || yi >= y)
20 20 return;
21 21  
22 22 // convert 2D coordinates to 1D
23   - size_t i = yi * x + xi;
  23 + int i = yi * x + xi;
24 24  
25 25 gpuCenters[i] = 0; //initialize the value at this location to zero
26 26  
27 27 T val = gpuVote[i];
28 28  
29   - //compare to the threshold
30   - //if(val < final_t) return;
31   -
32   - //define an array to store indices with same vote value
33   - /*int * IdxEq;
34   - IdxEq = new int [2*conn];
35   - int n = 0;*/
36   -
37 29 for(int xl = xi - conn; xl < xi + conn; xl++){
38 30 for(int yl = yi - conn; yl < yi + conn; yl++){
39 31 if(xl >= 0 && xl < x && yl >= 0 && yl < y){
... ... @@ -42,8 +34,7 @@ namespace stim{
42 34 return;
43 35 }
44 36 if (gpuVote[il] == val){
45   - /*IdxEq[n] = il;
46   - n = n+1;*/
  37 +
47 38 if( il > i){
48 39 return;
49 40 }
... ... @@ -51,12 +42,7 @@ namespace stim{
51 42 }
52 43 }
53 44 }
54   - /*if (n!=0){
55   - if(IdxEq[n/2] !=i){
56   - return;
57   - }
58   - } */
59   - //gpuCenters[i] = 1;
  45 +
60 46 gpuCenters[i] = gpuVote[i];
61 47 }
62 48  
... ...