Mirrors/overte
3
0
Fork 0
mirror of https://github.com/overte-org/overte.git synced 2025-04-18 22:16:28 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Merge pull request #8383 from AndrewMeadows/faster-is-better

Browse source 
faster image processing on load
This commit is contained in:
Brad Hefta-Gaub 2016-08-11 13:10:44 -07:00 committed by GitHub
commit ea1e090053
1 changed files with 22 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

48
libraries/model/src/model/TextureMap.cpp
View file

@ -59,31 +59,28 @@ const QImage TextureUsage::process2DImageColor(const QImage& srcImage, bool& val
const uint8 OPAQUE_ALPHA = 255;
const uint8 TRANSPARENT_ALPHA = 0;
if (image.hasAlphaChannel()) {
std::map<uint8, uint32> alphaHistogram;
if (image.format() != QImage::Format_ARGB32) {
image = image.convertToFormat(QImage::Format_ARGB32);
}
// Actual alpha channel? create the histogram
for (int y = 0; y < image.height(); ++y) {
const QRgb* data = reinterpret_cast<const QRgb*>(image.constScanLine(y));
for (int x = 0; x < image.width(); ++x) {
auto alpha = qAlpha(data[x]);
alphaHistogram[alpha] ++;
validAlpha = validAlpha || (alpha != OPAQUE_ALPHA);
// Figure out if we can use a mask for alpha or not
int numOpaques = 0;
int numTranslucents = 0;
const int NUM_PIXELS = image.width() * image.height();
const int MAX_TRANSLUCENT_PIXELS_FOR_ALPHAMASK = (int)(0.05f * (float)(NUM_PIXELS));
const QRgb* data = reinterpret_cast<const QRgb*>(image.constBits());
for (int i = 0; i < NUM_PIXELS; ++i) {
auto alpha = qAlpha(data[i]);
if (alpha == OPAQUE_ALPHA) {
numOpaques++;
} else if (alpha != TRANSPARENT_ALPHA) {
if (++numTranslucents > MAX_TRANSLUCENT_PIXELS_FOR_ALPHAMASK) {
alphaAsMask = false;
break;
}
}
}
// If alpha was meaningfull refine
if (validAlpha && (alphaHistogram.size() > 1)) {
auto totalNumPixels = image.height() * image.width();
auto numOpaques = alphaHistogram[OPAQUE_ALPHA];
auto numTransparents = alphaHistogram[TRANSPARENT_ALPHA];
auto numTranslucents = totalNumPixels - numOpaques - numTransparents;
alphaAsMask = ((numTranslucents / (double)totalNumPixels) < 0.05);
}
validAlpha = (numOpaques != NUM_PIXELS);
}
if (!validAlpha && image.format() != QImage::Format_RGB888) {
@ -660,13 +657,12 @@ const CubeLayout CubeLayout::CUBEMAP_LAYOUTS[] = {
const int CubeLayout::NUM_CUBEMAP_LAYOUTS = sizeof(CubeLayout::CUBEMAP_LAYOUTS) / sizeof(CubeLayout);
gpu::Texture* TextureUsage::processCubeTextureColorFromImage(const QImage& srcImage, const std::string& srcImageName, bool isLinear, bool doCompress, bool generateMips, bool generateIrradiance) {
bool validAlpha = false;
bool alphaAsMask = true;
QImage image = process2DImageColor(srcImage, validAlpha, alphaAsMask);
gpu::Texture* theTexture = nullptr;
if ((image.width() > 0) && (image.height() > 0)) {
if ((srcImage.width() > 0) && (srcImage.height() > 0)) {
QImage image = srcImage;
if (image.format() != QImage::Format_RGB888) {
image = image.convertToFormat(QImage::Format_RGB888);
}
gpu::Element formatGPU;
gpu::Element formatMip;
@ -674,7 +670,7 @@ gpu::Texture* TextureUsage::processCubeTextureColorFromImage(const QImage& srcIm
// Find the layout of the cubemap in the 2D image
int foundLayout = CubeLayout::findLayout(image.width(), image.height());
std::vector<QImage> faces;
// If found, go extract the faces as separate images
if (foundLayout >= 0) {