From 11c72b1132feca7c1252ea01d02da4cb497e723f Mon Sep 17 00:00:00 2001
From: Joseph Redmon <pjreddie@gmail.com>
Date: Thu, 11 Jun 2015 22:38:58 +0000
Subject: [PATCH] testing on one image
---
src/image.c | 836 +++++++++++++++++++++++++++++------------------------------
1 files changed, 409 insertions(+), 427 deletions(-)
diff --git a/src/image.c b/src/image.c
index 2cfce63..2e5d756 100644
--- a/src/image.c
+++ b/src/image.c
@@ -1,6 +1,12 @@
#include "image.h"
#include "utils.h"
#include <stdio.h>
+#include <math.h>
+
+#define STB_IMAGE_IMPLEMENTATION
+#include "stb_image.h"
+#define STB_IMAGE_WRITE_IMPLEMENTATION
+#include "stb_image_write.h"
int windows = 0;
@@ -19,7 +25,7 @@
void draw_box(image a, int x1, int y1, int x2, int y2, float r, float g, float b)
{
- normalize_image(a);
+ //normalize_image(a);
int i;
if(x1 < 0) x1 = 0;
if(x1 >= a.w) x1 = a.w-1;
@@ -53,21 +59,6 @@
}
}
-void jitter_image(image a, int h, int w, int dh, int dw)
-{
- int i,j,k;
- for(k = 0; k < a.c; ++k){
- for(i = 0; i < h; ++i){
- for(j = 0; j < w; ++j){
- int src = j + dw + (i+dh)*a.w + k*a.w*a.h;
- int dst = j + i*w + k*w*h;
- //printf("%d %d\n", src, dst);
- a.data[dst] = a.data[src];
- }
- }
- }
-}
-
void flip_image(image a)
{
int i,j,k;
@@ -87,7 +78,7 @@
image image_distance(image a, image b)
{
int i,j;
- image dist = make_image(a.h, a.w, 1);
+ image dist = make_image(a.w, a.h, 1);
for(i = 0; i < a.c; ++i){
for(j = 0; j < a.h*a.w; ++j){
dist.data[j] += pow(a.data[i*a.h*a.w+j]-b.data[i*a.h*a.w+j],2);
@@ -99,20 +90,14 @@
return dist;
}
-void subtract_image(image a, image b)
+void embed_image(image source, image dest, int dx, int dy)
{
- int i;
- for(i = 0; i < a.h*a.w*a.c; ++i) a.data[i] -= b.data[i];
-}
-
-void embed_image(image source, image dest, int h, int w)
-{
- int i,j,k;
+ int x,y,k;
for(k = 0; k < source.c; ++k){
- for(i = 0; i < source.h; ++i){
- for(j = 0; j < source.w; ++j){
- float val = get_pixel(source, i,j,k);
- set_pixel(dest, h+i, w+j, k, val);
+ for(y = 0; y < source.h; ++y){
+ for(x = 0; x < source.w; ++x){
+ float val = get_pixel(source, x,y,k);
+ set_pixel(dest, dx+x, dy+y, k, val);
}
}
}
@@ -122,20 +107,24 @@
{
int h = source.h;
h = (h+border)*source.c - border;
- image dest = make_image(h, source.w, 1);
+ image dest = make_image(source.w, h, 1);
int i;
for(i = 0; i < source.c; ++i){
image layer = get_image_layer(source, i);
int h_offset = i*(source.h+border);
- embed_image(layer, dest, h_offset, 0);
+ embed_image(layer, dest, 0, h_offset);
free_image(layer);
}
return dest;
}
-void z_normalize_image(image p)
+void constrain_image(image im)
{
- normalize_array(p.data, p.h*p.w*p.c);
+ int i;
+ for(i = 0; i < im.w*im.h*im.c; ++i){
+ if(im.data[i] < 0) im.data[i] = 0;
+ if(im.data[i] > 1) im.data[i] = 1;
+ }
}
void normalize_image(image p)
@@ -167,24 +156,6 @@
free(max);
}
-float avg_image_layer(image m, int l)
-{
- int i;
- float sum = 0;
- for(i = 0; i < m.h*m.w; ++i){
- sum += m.data[l*m.h*m.w + i];
- }
- return sum/(m.h*m.w);
-}
-
-void threshold_image(image p, float t)
-{
- int i;
- for(i = 0; i < p.w*p.h*p.c; ++i){
- if(p.data[i] < t) p.data[i] = 0;
- }
-}
-
image copy_image(image p)
{
image copy = p;
@@ -193,12 +164,23 @@
return copy;
}
-
-void show_image(image p, char *name)
+void rgbgr_image(image im)
{
- int i,j,k;
+ int i;
+ for(i = 0; i < im.w*im.h; ++i){
+ float swap = im.data[i];
+ im.data[i] = im.data[i+im.w*im.h*2];
+ im.data[i+im.w*im.h*2] = swap;
+ }
+}
+
+#ifdef OPENCV
+void show_image_cv(image p, char *name)
+{
+ int x,y,k;
image copy = copy_image(p);
- normalize_image(copy);
+ rgbgr_image(copy);
+ //normalize_image(copy);
char buff[256];
//sprintf(buff, "%s (%d)", name, windows);
@@ -209,16 +191,16 @@
cvNamedWindow(buff, CV_WINDOW_AUTOSIZE);
//cvMoveWindow(buff, 100*(windows%10) + 200*(windows/10), 100*(windows%10));
++windows;
- for(i = 0; i < p.h; ++i){
- for(j = 0; j < p.w; ++j){
+ for(y = 0; y < p.h; ++y){
+ for(x = 0; x < p.w; ++x){
for(k= 0; k < p.c; ++k){
- disp->imageData[i*step + j*p.c + k] = (unsigned char)(get_pixel(copy,i,j,k)*255);
+ disp->imageData[y*step + x*p.c + k] = (unsigned char)(get_pixel(copy,x,y,k)*255);
}
}
}
free_image(copy);
- if(disp->height < 500 || disp->width < 500 || disp->height > 1000){
- int w = 500;
+ if(disp->height < 448 || disp->width < 448 || disp->height > 1000){
+ int w = 448;
int h = w*p.h/p.w;
if(h > 1000){
h = 1000;
@@ -226,18 +208,46 @@
}
IplImage *buffer = disp;
disp = cvCreateImage(cvSize(w, h), buffer->depth, buffer->nChannels);
- cvResize(buffer, disp, CV_INTER_NN);
+ cvResize(buffer, disp, CV_INTER_LINEAR);
cvReleaseImage(&buffer);
}
cvShowImage(buff, disp);
cvReleaseImage(&disp);
}
+#endif
-void save_image(image p, char *name)
+void show_image(image p, char *name)
{
- int i,j,k;
+ #ifdef OPENCV
+ show_image_cv(p, name);
+ #else
+ fprintf(stderr, "Not compiled with OpenCV, saving to %s.png instead\n", name);
+ save_image(p, name);
+ #endif
+}
+
+void save_image(image im, char *name)
+{
+ char buff[256];
+ //sprintf(buff, "%s (%d)", name, windows);
+ sprintf(buff, "%s.png", name);
+ unsigned char *data = calloc(im.w*im.h*im.c, sizeof(char));
+ int i,k;
+ for(k = 0; k < im.c; ++k){
+ for(i = 0; i < im.w*im.h; ++i){
+ data[i*im.c+k] = (unsigned char) (255*im.data[i + k*im.w*im.h]);
+ }
+ }
+ int success = stbi_write_png(buff, im.w, im.h, im.c, data, im.w*im.c);
+ if(!success) fprintf(stderr, "Failed to write image %s\n", buff);
+}
+
+/*
+void save_image_cv(image p, char *name)
+{
+ int x,y,k;
image copy = copy_image(p);
- normalize_image(copy);
+ //normalize_image(copy);
char buff[256];
//sprintf(buff, "%s (%d)", name, windows);
@@ -245,10 +255,10 @@
IplImage *disp = cvCreateImage(cvSize(p.w,p.h), IPL_DEPTH_8U, p.c);
int step = disp->widthStep;
- for(i = 0; i < p.h; ++i){
- for(j = 0; j < p.w; ++j){
+ for(y = 0; y < p.h; ++y){
+ for(x = 0; x < p.w; ++x){
for(k= 0; k < p.c; ++k){
- disp->imageData[i*step + j*p.c + k] = (unsigned char)(get_pixel(copy,i,j,k)*255);
+ disp->imageData[y*step + x*p.c + k] = (unsigned char)(get_pixel(copy,x,y,k)*255);
}
}
}
@@ -256,6 +266,7 @@
cvSaveImage(buff, disp,0);
cvReleaseImage(&disp);
}
+*/
void show_image_layers(image p, char *name)
{
@@ -276,7 +287,7 @@
free_image(c);
}
-image make_empty_image(int h, int w, int c)
+image make_empty_image(int w, int h, int c)
{
image out;
out.data = 0;
@@ -286,63 +297,37 @@
return out;
}
-image make_image(int h, int w, int c)
+image make_image(int w, int h, int c)
{
- image out = make_empty_image(h,w,c);
+ image out = make_empty_image(w,h,c);
out.data = calloc(h*w*c, sizeof(float));
return out;
}
-image float_to_image(int h, int w, int c, float *data)
+image float_to_image(int w, int h, int c, float *data)
{
- image out = make_empty_image(h,w,c);
+ image out = make_empty_image(w,h,c);
out.data = data;
return out;
}
-void zero_image(image m)
+image rotate_image(image im, float rad)
{
- memset(m.data, 0, m.h*m.w*m.c*sizeof(float));
-}
-
-void zero_channel(image m, int c)
-{
- memset(&(m.data[c*m.h*m.w]), 0, m.h*m.w*sizeof(float));
-}
-
-void rotate_image(image m)
-{
- int i,j;
- for(j = 0; j < m.c; ++j){
- for(i = 0; i < m.h*m.w/2; ++i){
- float swap = m.data[j*m.h*m.w + i];
- m.data[j*m.h*m.w + i] = m.data[j*m.h*m.w + (m.h*m.w-1 - i)];
- m.data[j*m.h*m.w + (m.h*m.w-1 - i)] = swap;
- }
- }
-}
-
-image make_random_image(int h, int w, int c)
-{
- image out = make_image(h,w,c);
- int i;
- for(i = 0; i < h*w*c; ++i){
- out.data[i] = rand_normal();
- //out.data[i] = rand()%3;
- }
- return out;
-}
-
-void add_into_image(image src, image dest, int h, int w)
-{
- int i,j,k;
- for(k = 0; k < src.c; ++k){
- for(i = 0; i < src.h; ++i){
- for(j = 0; j < src.w; ++j){
- add_pixel(dest, h+i, w+j, k, get_pixel(src, i, j, k));
+ int x, y, c;
+ float cx = im.w/2.;
+ float cy = im.h/2.;
+ image rot = make_image(im.w, im.h, im.c);
+ for(c = 0; c < im.c; ++c){
+ for(y = 0; y < im.h; ++y){
+ for(x = 0; x < im.w; ++x){
+ float rx = cos(rad)*(x-cx) - sin(rad)*(y-cy) + cx;
+ float ry = sin(rad)*(x-cx) + cos(rad)*(y-cy) + cy;
+ float val = billinear_interpolate(im, rx, ry, c);
+ set_pixel(rot, x, y, c, val);
}
}
}
+ return rot;
}
void translate_image(image m, float s)
@@ -357,85 +342,232 @@
for(i = 0; i < m.h*m.w*m.c; ++i) m.data[i] *= s;
}
-image make_random_kernel(int size, int c, float scale)
+image crop_image(image im, int dx, int dy, int w, int h)
{
- int pad;
- if((pad=(size%2==0))) ++size;
- image out = make_random_image(size,size,c);
- scale_image(out, scale);
- int i,k;
- if(pad){
- for(k = 0; k < out.c; ++k){
- for(i = 0; i < size; ++i) {
- set_pixel(out, i, 0, k, 0);
- set_pixel(out, 0, i, k, 0);
- }
- }
- }
- return out;
-}
-
-image ipl_to_image(IplImage* src)
-{
- unsigned char *data = (unsigned char *)src->imageData;
- int h = src->height;
- int w = src->width;
- int c = src->nChannels;
- int step = src->widthStep;
- image out = make_image(h,w,c);
- int i, j, k, count=0;;
-
- for(k= 0; k < c; ++k){
- for(i = 0; i < h; ++i){
- for(j = 0; j < w; ++j){
- out.data[count++] = data[i*step + j*c + k];
- }
- }
- }
- return out;
-}
-
-image crop_image(image im, int dr, int dc, int h, int w)
-{
- image cropped = make_image(h, w, im.c);
+ image cropped = make_image(w, h, im.c);
int i, j, k;
for(k = 0; k < im.c; ++k){
for(j = 0; j < h; ++j){
for(i = 0; i < w; ++i){
- int r = j + dr;
- int c = i + dc;
- float val = 128;
+ int r = j + dy;
+ int c = i + dx;
+ float val = 0;
if (r >= 0 && r < im.h && c >= 0 && c < im.w) {
- val = get_pixel(im, r, c, k);
+ val = get_pixel(im, c, r, k);
}
- set_pixel(cropped, j, i, k, val);
+ set_pixel(cropped, i, j, k, val);
}
}
}
return cropped;
}
-// #wikipedia
-image resize_image(image im, int h, int w)
+float three_way_max(float a, float b, float c)
{
- image resized = make_image(h, w, im.c);
+ return (a > b) ? ( (a > c) ? a : c) : ( (b > c) ? b : c) ;
+}
+
+float three_way_min(float a, float b, float c)
+{
+ return (a < b) ? ( (a < c) ? a : c) : ( (b < c) ? b : c) ;
+}
+
+// http://www.cs.rit.edu/~ncs/color/t_convert.html
+void rgb_to_hsv(image im)
+{
+ assert(im.c == 3);
+ int i, j;
+ float r, g, b;
+ float h, s, v;
+ for(j = 0; j < im.h; ++j){
+ for(i = 0; i < im.w; ++i){
+ r = get_pixel(im, i , j, 0);
+ g = get_pixel(im, i , j, 1);
+ b = get_pixel(im, i , j, 2);
+ float max = three_way_max(r,g,b);
+ float min = three_way_min(r,g,b);
+ float delta = max - min;
+ v = max;
+ if(max == 0){
+ s = 0;
+ h = -1;
+ }else{
+ s = delta/max;
+ if(r == max){
+ h = (g - b) / delta;
+ } else if (g == max) {
+ h = 2 + (b - r) / delta;
+ } else {
+ h = 4 + (r - g) / delta;
+ }
+ if (h < 0) h += 6;
+ }
+ set_pixel(im, i, j, 0, h);
+ set_pixel(im, i, j, 1, s);
+ set_pixel(im, i, j, 2, v);
+ }
+ }
+}
+
+void hsv_to_rgb(image im)
+{
+ assert(im.c == 3);
+ int i, j;
+ float r, g, b;
+ float h, s, v;
+ float f, p, q, t;
+ for(j = 0; j < im.h; ++j){
+ for(i = 0; i < im.w; ++i){
+ h = get_pixel(im, i , j, 0);
+ s = get_pixel(im, i , j, 1);
+ v = get_pixel(im, i , j, 2);
+ if (s == 0) {
+ r = g = b = v;
+ } else {
+ int index = floor(h);
+ f = h - index;
+ p = v*(1-s);
+ q = v*(1-s*f);
+ t = v*(1-s*(1-f));
+ if(index == 0){
+ r = v; g = t; b = p;
+ } else if(index == 1){
+ r = q; g = v; b = p;
+ } else if(index == 2){
+ r = p; g = v; b = t;
+ } else if(index == 3){
+ r = p; g = q; b = v;
+ } else if(index == 4){
+ r = t; g = p; b = v;
+ } else {
+ r = v; g = p; b = q;
+ }
+ }
+ set_pixel(im, i, j, 0, r);
+ set_pixel(im, i, j, 1, g);
+ set_pixel(im, i, j, 2, b);
+ }
+ }
+}
+
+image grayscale_image(image im)
+{
+ assert(im.c == 3);
+ int i, j, k;
+ image gray = make_image(im.w, im.h, im.c);
+ float scale[] = {0.587, 0.299, 0.114};
+ for(k = 0; k < im.c; ++k){
+ for(j = 0; j < im.h; ++j){
+ for(i = 0; i < im.w; ++i){
+ gray.data[i+im.w*j] += scale[k]*get_pixel(im, i, j, k);
+ }
+ }
+ }
+ memcpy(gray.data + im.w*im.h*1, gray.data, sizeof(float)*im.w*im.h);
+ memcpy(gray.data + im.w*im.h*2, gray.data, sizeof(float)*im.w*im.h);
+ return gray;
+}
+
+image blend_image(image fore, image back, float alpha)
+{
+ assert(fore.w == back.w && fore.h == back.h && fore.c == back.c);
+ image blend = make_image(fore.w, fore.h, fore.c);
+ int i, j, k;
+ for(k = 0; k < fore.c; ++k){
+ for(j = 0; j < fore.h; ++j){
+ for(i = 0; i < fore.w; ++i){
+ float val = alpha * get_pixel(fore, i, j, k) +
+ (1 - alpha)* get_pixel(back, i, j, k);
+ set_pixel(blend, i, j, k, val);
+ }
+ }
+ }
+ return blend;
+}
+
+void scale_image_channel(image im, int c, float v)
+{
+ int i, j;
+ for(j = 0; j < im.h; ++j){
+ for(i = 0; i < im.w; ++i){
+ float pix = get_pixel(im, i, j, c);
+ pix = pix*v;
+ set_pixel(im, i, j, c, pix);
+ }
+ }
+}
+
+void saturate_image(image im, float sat)
+{
+ rgb_to_hsv(im);
+ scale_image_channel(im, 1, sat);
+ hsv_to_rgb(im);
+ constrain_image(im);
+}
+
+void exposure_image(image im, float sat)
+{
+ rgb_to_hsv(im);
+ scale_image_channel(im, 2, sat);
+ hsv_to_rgb(im);
+ constrain_image(im);
+}
+
+void saturate_exposure_image(image im, float sat, float exposure)
+{
+ rgb_to_hsv(im);
+ scale_image_channel(im, 1, sat);
+ scale_image_channel(im, 2, exposure);
+ hsv_to_rgb(im);
+ constrain_image(im);
+}
+
+/*
+ image saturate_image(image im, float sat)
+ {
+ image gray = grayscale_image(im);
+ image blend = blend_image(im, gray, sat);
+ free_image(gray);
+ constrain_image(blend);
+ return blend;
+ }
+
+ image brightness_image(image im, float b)
+ {
+ image bright = make_image(im.w, im.h, im.c);
+ return bright;
+ }
+ */
+
+float billinear_interpolate(image im, float x, float y, int c)
+{
+ int ix = (int) floorf(x);
+ int iy = (int) floorf(y);
+
+ float dx = x - ix;
+ float dy = y - iy;
+
+ float val = (1-dy) * (1-dx) * get_pixel_extend(im, ix, iy, c) +
+ dy * (1-dx) * get_pixel_extend(im, ix, iy+1, c) +
+ (1-dy) * dx * get_pixel_extend(im, ix+1, iy, c) +
+ dy * dx * get_pixel_extend(im, ix+1, iy+1, c);
+ return val;
+}
+
+// #wikipedia
+image resize_image(image im, int w, int h)
+{
+ image resized = make_image(w, h, im.c);
int r, c, k;
- float h_scale = (float)(im.h - 1) / (h - 1) - .00001;
- float w_scale = (float)(im.w - 1) / (w - 1) - .00001;
+ float w_scale = (float)(im.w - 1) / (w - 1);
+ float h_scale = (float)(im.h - 1) / (h - 1);
for(k = 0; k < im.c; ++k){
for(r = 0; r < h; ++r){
for(c = 0; c < w; ++c){
- float sr = r*h_scale;
- float sc = c*w_scale;
- int ir = (int)sr;
- int ic = (int)sc;
- float x = sr-ir;
- float y = sc-ic;
- float val = (1-x) * (1-y) * get_pixel(im, ir, ic, k) +
- x * (1-y) * get_pixel(im, ir+1, ic, k) +
- (1-x) * y * get_pixel(im, ir, ic+1, k) +
- x * y * get_pixel(im, ir+1, ic+1, k);
- set_pixel(resized, r, c, k, val);
+ float sx = c*w_scale;
+ float sy = r*h_scale;
+ float val = billinear_interpolate(im, sx, sy, k);
+ set_pixel(resized, c, r, k, val);
}
}
}
@@ -444,260 +576,147 @@
void test_resize(char *filename)
{
- image im = load_image(filename, 0,0);
- image small = resize_image(im, 63, 65);
- image big = resize_image(im, 512, 513);
- image crop = crop_image(im, 10, 50, 100, 100);
- image crop2 = crop_image(im, -50, -30, 400, 291);
- show_image(im, "original");
- show_image(small, "smaller");
- show_image(big, "bigger");
- show_image(crop, "crop");
- show_image(crop2, "crop2");
+ image im = load_image(filename, 0,0, 3);
+ image gray = grayscale_image(im);
+
+ image sat2 = copy_image(im);
+ saturate_image(sat2, 2);
+
+ image sat5 = copy_image(im);
+ saturate_image(sat5, .5);
+
+ image exp2 = copy_image(im);
+ exposure_image(exp2, 2);
+
+ image exp5 = copy_image(im);
+ exposure_image(exp5, .5);
+
+ show_image(im, "Original");
+ show_image(gray, "Gray");
+ show_image(sat2, "Saturation-2");
+ show_image(sat5, "Saturation-.5");
+ show_image(exp2, "Exposure-2");
+ show_image(exp5, "Exposure-.5");
+ #ifdef OPENCV
cvWaitKey(0);
+ #endif
}
-image load_image_color(char *filename, int h, int w)
+#ifdef OPENCV
+image ipl_to_image(IplImage* src)
+{
+ unsigned char *data = (unsigned char *)src->imageData;
+ int h = src->height;
+ int w = src->width;
+ int c = src->nChannels;
+ int step = src->widthStep;
+ image out = make_image(w, h, c);
+ int i, j, k, count=0;;
+
+ for(k= 0; k < c; ++k){
+ for(i = 0; i < h; ++i){
+ for(j = 0; j < w; ++j){
+ out.data[count++] = data[i*step + j*c + k]/255.;
+ }
+ }
+ }
+ return out;
+}
+
+image load_image_cv(char *filename, int channels)
{
IplImage* src = 0;
- if( (src = cvLoadImage(filename, 1)) == 0 )
+ int flag = -1;
+ if (channels == 0) flag = -1;
+ else if (channels == 1) flag = 0;
+ else if (channels == 3) flag = 1;
+ else {
+ fprintf(stderr, "OpenCV can't force load with %d channels\n", channels);
+ }
+
+ if( (src = cvLoadImage(filename, flag)) == 0 )
{
printf("Cannot load file image %s\n", filename);
exit(0);
}
image out = ipl_to_image(src);
- if((h && w) && (h != out.h || w != out.w)){
- image resized = resize_image(out, h, w);
- free_image(out);
- out = resized;
- }
cvReleaseImage(&src);
+ rgbgr_image(out);
return out;
}
-image load_image(char *filename, int h, int w)
+#endif
+
+
+image load_image_stb(char *filename, int channels)
{
- IplImage* src = 0;
- if( (src = cvLoadImage(filename,-1)) == 0 )
- {
- printf("Cannot load file image %s\n", filename);
+ int w, h, c;
+ unsigned char *data = stbi_load(filename, &w, &h, &c, channels);
+ if (!data) {
+ fprintf(stderr, "Cannot load file image %s\nSTB Reason: %s\n", filename, stbi_failure_reason());
exit(0);
}
- image out = ipl_to_image(src);
+ if(channels) c = channels;
+ int i,j,k;
+ image im = make_image(w, h, c);
+ for(k = 0; k < c; ++k){
+ for(j = 0; j < h; ++j){
+ for(i = 0; i < w; ++i){
+ int dst_index = i + w*j + w*h*k;
+ int src_index = k + c*i + c*w*j;
+ im.data[dst_index] = (float)data[src_index]/255.;
+ }
+ }
+ }
+ free(data);
+ return im;
+}
+
+image load_image(char *filename, int w, int h, int c)
+{
+ #ifdef OPENCV
+ image out = load_image_cv(filename, c);
+ #else
+ image out = load_image_stb(filename, c);
+ #endif
+
if((h && w) && (h != out.h || w != out.w)){
- image resized = resize_image(out, h, w);
+ image resized = resize_image(out, w, h);
free_image(out);
out = resized;
}
- cvReleaseImage(&src);
return out;
}
+image load_image_color(char *filename, int w, int h)
+{
+ return load_image(filename, w, h, 3);
+}
+
image get_image_layer(image m, int l)
{
- image out = make_image(m.h, m.w, 1);
+ image out = make_image(m.w, m.h, 1);
int i;
for(i = 0; i < m.h*m.w; ++i){
out.data[i] = m.data[i+l*m.h*m.w];
}
return out;
}
-image get_sub_image(image m, int h, int w, int dh, int dw)
-{
- image out = make_image(dh, dw, m.c);
- int i,j,k;
- for(k = 0; k < out.c; ++k){
- for(i = 0; i < dh; ++i){
- for(j = 0; j < dw; ++j){
- float val = get_pixel(m, h+i, w+j, k);
- set_pixel(out, i, j, k, val);
- }
- }
- }
- return out;
-}
float get_pixel(image m, int x, int y, int c)
{
- assert(x < m.h && y < m.w && c < m.c);
- return m.data[c*m.h*m.w + x*m.w + y];
+ assert(x < m.w && y < m.h && c < m.c);
+ return m.data[c*m.h*m.w + y*m.w + x];
}
float get_pixel_extend(image m, int x, int y, int c)
{
- if(x < 0 || x >= m.h || y < 0 || y >= m.w || c < 0 || c >= m.c) return 0;
+ if(x < 0 || x >= m.w || y < 0 || y >= m.h || c < 0 || c >= m.c) return 0;
return get_pixel(m, x, y, c);
}
void set_pixel(image m, int x, int y, int c, float val)
{
- assert(x < m.h && y < m.w && c < m.c);
- m.data[c*m.h*m.w + x*m.w + y] = val;
-}
-void set_pixel_extend(image m, int x, int y, int c, float val)
-{
- if(x < 0 || x >= m.h || y < 0 || y >= m.w || c < 0 || c >= m.c) return;
- set_pixel(m, x, y, c, val);
-}
-
-void add_pixel(image m, int x, int y, int c, float val)
-{
- assert(x < m.h && y < m.w && c < m.c);
- m.data[c*m.h*m.w + x*m.w + y] += val;
-}
-
-void add_pixel_extend(image m, int x, int y, int c, float val)
-{
- if(x < 0 || x >= m.h || y < 0 || y >= m.w || c < 0 || c >= m.c) return;
- add_pixel(m, x, y, c, val);
-}
-
-void two_d_convolve(image m, int mc, image kernel, int kc, int stride, image out, int oc, int edge)
-{
- int x,y,i,j;
- int xstart, xend, ystart, yend;
- if(edge){
- xstart = ystart = 0;
- xend = m.h;
- yend = m.w;
- }else{
- xstart = kernel.h/2;
- ystart = kernel.w/2;
- xend = m.h-kernel.h/2;
- yend = m.w - kernel.w/2;
- }
- for(x = xstart; x < xend; x += stride){
- for(y = ystart; y < yend; y += stride){
- float sum = 0;
- for(i = 0; i < kernel.h; ++i){
- for(j = 0; j < kernel.w; ++j){
- sum += get_pixel(kernel, i, j, kc)*get_pixel_extend(m, x+i-kernel.h/2, y+j-kernel.w/2, mc);
- }
- }
- add_pixel(out, (x-xstart)/stride, (y-ystart)/stride, oc, sum);
- }
- }
-}
-
-float single_convolve(image m, image kernel, int x, int y)
-{
- float sum = 0;
- int i, j, k;
- for(i = 0; i < kernel.h; ++i){
- for(j = 0; j < kernel.w; ++j){
- for(k = 0; k < kernel.c; ++k){
- sum += get_pixel(kernel, i, j, k)*get_pixel_extend(m, x+i-kernel.h/2, y+j-kernel.w/2, k);
- }
- }
- }
- return sum;
-}
-
-void convolve(image m, image kernel, int stride, int channel, image out, int edge)
-{
- assert(m.c == kernel.c);
- int i;
- zero_channel(out, channel);
- for(i = 0; i < m.c; ++i){
- two_d_convolve(m, i, kernel, i, stride, out, channel, edge);
- }
- /*
- int j;
- for(i = 0; i < m.h; i += stride){
- for(j = 0; j < m.w; j += stride){
- float val = single_convolve(m, kernel, i, j);
- set_pixel(out, i/stride, j/stride, channel, val);
- }
- }
- */
-}
-
-void upsample_image(image m, int stride, image out)
-{
- int i,j,k;
- zero_image(out);
- for(k = 0; k < m.c; ++k){
- for(i = 0; i < m.h; ++i){
- for(j = 0; j< m.w; ++j){
- float val = get_pixel(m, i, j, k);
- set_pixel(out, i*stride, j*stride, k, val);
- }
- }
- }
-}
-
-void single_update(image m, image update, int x, int y, float error)
-{
- int i, j, k;
- for(i = 0; i < update.h; ++i){
- for(j = 0; j < update.w; ++j){
- for(k = 0; k < update.c; ++k){
- float val = get_pixel_extend(m, x+i-update.h/2, y+j-update.w/2, k);
- add_pixel(update, i, j, k, val*error);
- }
- }
- }
-}
-
-void kernel_update(image m, image update, int stride, int channel, image out, int edge)
-{
- assert(m.c == update.c);
- zero_image(update);
- int i, j, istart, jstart, iend, jend;
- if(edge){
- istart = jstart = 0;
- iend = m.h;
- jend = m.w;
- }else{
- istart = update.h/2;
- jstart = update.w/2;
- iend = m.h-update.h/2;
- jend = m.w - update.w/2;
- }
- for(i = istart; i < iend; i += stride){
- for(j = jstart; j < jend; j += stride){
- float error = get_pixel(out, (i-istart)/stride, (j-jstart)/stride, channel);
- single_update(m, update, i, j, error);
- }
- }
- /*
- for(i = 0; i < update.h*update.w*update.c; ++i){
- update.data[i] /= (m.h/stride)*(m.w/stride);
- }
- */
-}
-
-void single_back_convolve(image m, image kernel, int x, int y, float val)
-{
- int i, j, k;
- for(i = 0; i < kernel.h; ++i){
- for(j = 0; j < kernel.w; ++j){
- for(k = 0; k < kernel.c; ++k){
- float pval = get_pixel(kernel, i, j, k) * val;
- add_pixel_extend(m, x+i-kernel.h/2, y+j-kernel.w/2, k, pval);
- }
- }
- }
-}
-
-void back_convolve(image m, image kernel, int stride, int channel, image out, int edge)
-{
- assert(m.c == kernel.c);
- int i, j, istart, jstart, iend, jend;
- if(edge){
- istart = jstart = 0;
- iend = m.h;
- jend = m.w;
- }else{
- istart = kernel.h/2;
- jstart = kernel.w/2;
- iend = m.h-kernel.h/2;
- jend = m.w - kernel.w/2;
- }
- for(i = istart; i < iend; i += stride){
- for(j = jstart; j < jend; j += stride){
- float val = get_pixel(out, (i-istart)/stride, (j-jstart)/stride, channel);
- single_back_convolve(m, kernel, i, j, val);
- }
- }
+ assert(x < m.w && y < m.h && c < m.c);
+ m.data[c*m.h*m.w + y*m.w + x] = val;
}
void print_image(image m)
@@ -730,20 +749,20 @@
c = 1;
}
- image filters = make_image(h,w,c);
+ image filters = make_image(w, h, c);
int i,j;
for(i = 0; i < n; ++i){
int h_offset = i*(ims[0].h+border);
image copy = copy_image(ims[i]);
//normalize_image(copy);
if(c == 3 && color){
- embed_image(copy, filters, h_offset, 0);
+ embed_image(copy, filters, 0, h_offset);
}
else{
for(j = 0; j < copy.c; ++j){
int w_offset = j*(ims[0].w+border);
image layer = get_image_layer(copy, j);
- embed_image(layer, filters, h_offset, w_offset);
+ embed_image(layer, filters, w_offset, h_offset);
free_image(layer);
}
}
@@ -766,20 +785,20 @@
c = 1;
}
- image filters = make_image(h,w,c);
+ image filters = make_image(w, h, c);
int i,j;
for(i = 0; i < n; ++i){
int w_offset = i*(size+border);
image copy = copy_image(ims[i]);
//normalize_image(copy);
if(c == 3 && color){
- embed_image(copy, filters, 0, w_offset);
+ embed_image(copy, filters, w_offset, 0);
}
else{
for(j = 0; j < copy.c; ++j){
int h_offset = j*(size+border);
image layer = get_image_layer(copy, j);
- embed_image(layer, filters, h_offset, w_offset);
+ embed_image(layer, filters, w_offset, h_offset);
free_image(layer);
}
}
@@ -796,43 +815,6 @@
free_image(m);
}
-image grid_images(image **ims, int h, int w)
-{
- int i;
- image *rows = calloc(h, sizeof(image));
- for(i = 0; i < h; ++i){
- rows[i] = collapse_images_horz(ims[i], w);
- }
- image out = collapse_images_vert(rows, h);
- for(i = 0; i < h; ++i){
- free_image(rows[i]);
- }
- free(rows);
- return out;
-}
-
-void test_grid()
-{
- int i,j;
- int num = 3;
- int topk = 3;
- image **vizs = calloc(num, sizeof(image*));
- for(i = 0; i < num; ++i){
- vizs[i] = calloc(topk, sizeof(image));
- for(j = 0; j < topk; ++j) vizs[i][j] = make_image(3,3,3);
- }
- image grid = grid_images(vizs, num, topk);
- save_image(grid, "Test Grid");
- free_image(grid);
-}
-
-void show_images_grid(image **ims, int h, int w, char *window)
-{
- image out = grid_images(ims, h, w);
- show_image(out, window);
- free_image(out);
-}
-
void free_image(image m)
{
free(m.data);
--
Gitblit v1.10.0