A problem with the form of converting ply to fbx - assimp

We are currently in the process of converting PLY or PCD data to FBX.
I did the conversion using the Assimp library as shown below, but I found a problem with missing colors.
May I know where is wrong?
#include <assimp/Importer.hpp>
#include <assimp/scene.h>
#include <assimp/postprocess.h>
#include <assimp/Exporter.hpp>
#include <System.h>
int main(int argc, char* argv[]) {
Assimp::Importer importer;
const aiScene* scene = importer.ReadFile("/app/test.ply",
aiProcess_Triangulate |
aiProcess_JoinIdenticalVertices |
aiProcess_SortByPType);
// Check that the import was successful
if (!scene) {
std::cerr << "Failed to import file: " << importer.GetErrorString() << std::endl;
return 1;
}
// Iterate over all the meshes in the scene
for (unsigned int i = 0; i < scene->mNumMeshes; ++i) {
aiMesh* mesh = scene->mMeshes[i];
// Check if the mesh has color data
if (mesh->HasVertexColors(0)) {
// Set the number of color components to 3 (for RGB)
//mesh->mNumColorComponents = 3;
// Allocate memory for the color data
mesh->mColors[0] = new aiColor4D[mesh->mNumVertices];
// Fill in the color data (for example, set all vertices to red)
for (unsigned int j = 0; j < mesh->mNumVertices; ++j) {
mesh->mColors[0][j].r = 1.0f;
mesh->mColors[0][j].g = 0.0f;
mesh->mColors[0][j].b = 0.0f;
mesh->mColors[0][j].a = 1.0f;
}
}
}
// Export the scene to fbx format
Assimp::Exporter exporter;
exporter.Export(scene, "fbx", "output.fbx");
return 0;
}

Related

Segmentation fault is thrown when using Metal (Mac OS X 10.15.6)

I am trying to learn Metal through the Apple documentation. So far, I have finished writing an application that calculates the square root of 4096 random numbers. However, when I run it through the terminal, it immediately throws a segmentation fault.
Output:
Segmentation fault: 11
logout
Saving session...
...copying shared history...
...saving history...truncating history files...
...completed.
[Process completed]
So far, I have tried inserting std::couts almost everywhere in the code and I have found the problem to be with the function that generates the random numbers (generateRandomFloatData(id<MTLBuffer> buffer)).
When I tried to print out the address of the input buffer, I got this output:
0x0
Segmentation fault: 11
logout
Saving session...
...copying shared history...
...saving history...truncating history files...
...completed.
[Process completed]
Weirdly, it prints out the address of a NULL pointer.
More testing revealed that changing the function to input a char pointer correctly outputs an address 0x7ffee8bd8620 pointing to the string.
Is there a problem in my code?
//
// main.mm
// MetalComputeCPP
//
// Created by [] on 5/1/21.
// Copyright © 2021 thng. All rights reserved.
//
#include <iostream>
#include <ApplicationServices/ApplicationServices.h>
#include <Metal/Metal.h>
#include <Foundation/Foundation.h>
#include <chrono>
const unsigned int arrayLength = 1 << 12;
const unsigned int bufferSize = arrayLength * sizeof(float);
void generateRandomFloatData(id<MTLBuffer> buffer) {
std::cout << ((float*)buffer.contents) << "\n";
float* dataPtr = ((float*)buffer.contents);
for (unsigned long index = 0; index < arrayLength; index++)
{
dataPtr[index] = (float)((rand()/(float)(RAND_MAX))*10);
std::cout << dataPtr[index] << "\n";
}
}
int main(int argc, const char * argv[]) {
id<MTLDevice> _mDevice = MTLCreateSystemDefaultDevice();
NSError* error = nil;
id<MTLLibrary> defaultLibrary = [_mDevice newDefaultLibrary];
id<MTLFunction> SqrtFunction = [defaultLibrary newFunctionWithName:#"SqrtArray"];
id<MTLComputePipelineState> _mSqrtFunctionPSO = [_mDevice newComputePipelineStateWithFunction: SqrtFunction error:&error];
id<MTLCommandQueue> _mCommandQueue = _mDevice.newCommandQueue;
id<MTLBuffer> _mBufferA = [_mDevice newBufferWithLength:bufferSize options:MTLResourceStorageModeShared];
id<MTLBuffer> _mBufferResult = [_mDevice newBufferWithLength:bufferSize options:MTLResourceStorageModeShared];
MTLSize gridSize = MTLSizeMake(arrayLength, 1, 1);
NSUInteger threadGroupSize = _mSqrtFunctionPSO.maxTotalThreadsPerThreadgroup;
if (threadGroupSize > arrayLength)
{
threadGroupSize = arrayLength;
}
MTLSize threadgroupSize = MTLSizeMake(threadGroupSize, 1, 1);
generateRandomFloatData(_mBufferA);
std::cout << "Generated random float data.\n";
id<MTLCommandBuffer> commandBuffer = _mCommandQueue.commandBuffer;
id<MTLComputeCommandEncoder> computeEncoder = [commandBuffer computeCommandEncoder];
[computeEncoder setComputePipelineState:_mSqrtFunctionPSO];
[computeEncoder setBuffer:_mBufferA offset:0 atIndex:0];
[computeEncoder setBuffer:_mBufferResult offset:0 atIndex:1];
[computeEncoder dispatchThreads:gridSize
threadsPerThreadgroup:threadgroupSize];
[computeEncoder endEncoding];
[commandBuffer commit];
std::chrono::high_resolution_clock::time_point start = std::chrono::high_resolution_clock::now();
[commandBuffer waitUntilCompleted];
std::chrono::high_resolution_clock::time_point end = std::chrono::high_resolution_clock::now();
uint64_t time = std::chrono::duration_cast<std::chrono::nanoseconds>(end-start).count();
float* a = ((float*)_mBufferA.contents);
float* result = ((float*)_mBufferResult.contents);
bool err = false;
for (unsigned long index = 0; index < arrayLength; index++)
{
if (abs(result[index] - (float)sqrt(a[index])) > 0.0001) err = true;
std::cout << "√" << a[index] << (err ? " != " : " = ") << result[index] << "\n";
}
std::cout << time << " nanoseconds\n";
printf("Compute results as expected\n");
return 0;
}
//
// File.metal
// MetalComputeCPP
//
// Created by [] on 5/1/21.
// Copyright © 2021 thng. All rights reserved.
//
#include <metal_stdlib>
using namespace metal;
kernel void SqrtArray(device const float* inA,
device float* outB,
uint ind [[thread_position_in_grid]]) {
//(x^n-k)' = (nx^(n-1))
//f(x0)/f'(x0)
outB[ind] = 0.1;
for (int i = 0; i < 20; i++) {
outB[ind] = outB[ind]-((outB[ind]*outB[ind]-inA[ind])/(outB[ind]*2));
}
}
buffer in generateRandomFloatData is nil because _mBufferA is nil.
_mBufferA is nil because _mDevice is nil.
MTLCreateSystemDefaultDevice returns nil because (from MTLCreateSystemDefaultDevice)
In macOS, in order for the system to provide a default Metal device object, you must link to the CoreGraphics framework. You usually need to do this explicitly if you are writing apps that don't use graphics by default, such as command line tools.
Your previous question:
Why does Metal not work when run via the Terminal but is fine when run through Xcode?
In Xcode MTLCreateSystemDefaultDevice returns on my Mac
_mDevice: <CaptureMTLDevice: 0x10050bbb0> -> <MTLDebugDevice: 0x10050aae0> -> <MTLIGAccelDevice: 0x1031c8000>
name = Intel HD Graphics 4000
In Terminal MTLCreateSystemDefaultDevice returns
_mDevice: <MTLIGAccelDevice: 0x7f9c32f17000>
name = Intel HD Graphics 4000
Apparenlty Xcode wraps the device in a debugging device, which has the side effect of fixing the issue.

Qimages container

I’m looking for a class like QSprite, which takes responsibility for reading pixmaps from a source file, split the source image into sequence of images and return sub image by index.
Something like this:
class QSprite2
{
bool load(QUrl url, QSize frameSize);
QImage getFrame(int index);
};
Is there any way to get described functionality with existing classes. Or maybe I have to implement the describe logic by myself?
You can use QQuickImageProvider. Here's a class I wrote that does something similar:
SpriteImageProvider.h:
#ifndef SPRITEIMAGEPROVIDER_H
#define SPRITEIMAGEPROVIDER_H
#include <QHash>
#include <QImage>
#include <QString>
#include <QQuickImageProvider>
#include "IsleGlobal.h"
class ISLE_EXPORT SpriteImageProvider : public QQuickImageProvider
{
public:
SpriteImageProvider();
QImage requestImage(const QString &id, QSize *size, const QSize &requestedSize) override;
private:
int parseFrameIndex(const QString &frameIndexStr) const;
int parseFrameWidth(const QString &frameWidthStr) const;
int parseFrameHeight(const QString &frameHeightStr) const;
QHash<QString, QImage> mImages;
};
#endif // SPRITEIMAGEPROVIDER_H
SpriteImageProvider.cpp:
#include "SpriteImageProvider.h"
#include <QImage>
#include <QDebug>
SpriteImageProvider::SpriteImageProvider() :
QQuickImageProvider(QQmlImageProviderBase::Image)
{
}
static const QRgb whiteRgba = qRgba(255, 255, 255, 255);
static const QRgb magentaRgba = qRgba(255, 0, 255, 255);
static const QRgb transparentRgba = qRgba(0, 0, 0, 0);
QImage SpriteImageProvider::requestImage(const QString &id, QSize *size, const QSize &)
{
QStringList args = id.split(QLatin1String(","));
if (args.length() < 1) {
qWarning() << "Must pass at least the file name as arguments for SpriteImageProvider source";
return QImage();
} else if (args.length() != 1 && args.length() != 2 && args.length() != 4) {
qWarning() << "Must pass either (fileName), (fileName, frameIndex) or"
<< "(fileName, frameWidth, frameHeight, frameIndex) as arguments for SpriteImageProvider source";
return QImage();
}
const QString imageFilename = ":/" + args.first();
int frameIndex = -2;
int frameWidth = -2;
int frameHeight = -2;
if (args.length() > 1) {
frameIndex = parseFrameIndex(args.last());
if (frameIndex == -1)
return QImage();
}
if (args.length() == 4) {
frameWidth = parseFrameIndex(args.at(1));
if (frameWidth == -1)
return QImage();
frameHeight = parseFrameIndex(args.at(2));
if (frameHeight == -1)
return QImage();
}
QHash<QString, QImage>::const_iterator it = mImages.find(imageFilename);
if (it == mImages.end()) {
QImage image(imageFilename);
if (image.isNull()) {
qWarning() << "Failed to load image at" << imageFilename;
return image;
}
// TODO: is there a better way of doing this?
QImage alphaImage = image;
for (int y = 0; y < alphaImage.height(); ++y) {
for (int x = 0; x < alphaImage.width(); ++x) {
const QRgb pixelRgb = alphaImage.pixel(x, y);
if (pixelRgb == whiteRgba || pixelRgb == magentaRgba)
alphaImage.setPixel(x, y, transparentRgba);
}
}
mImages.insert(imageFilename, alphaImage);
it = mImages.find(imageFilename);
}
if (frameWidth == -2 || frameHeight == -2) {
if (frameIndex == -2) {
// Use the whole image.
frameWidth = it.value().width();
frameHeight = it.value().height();
frameIndex = 0;
} else {
frameWidth = 64;
frameHeight = 64;
}
}
// Copy an individual frame out of the larger image.
const int framesWide = it.value().width() / frameWidth;
const QRect subRect((frameIndex % framesWide) * frameWidth,
(frameIndex / framesWide) * frameHeight,
frameWidth,
frameHeight);
// qDebug() << "id" << id;
// qDebug() << "framesWide" << framesWide;
// qDebug() << "subRect" << subRect;
const QImage frameImage = it.value().copy(subRect);
*size = frameImage.size();
return frameImage;
}
int SpriteImageProvider::parseFrameIndex(const QString &frameIndexStr) const
{
bool convertedToIntSuccessfully = false;
const int frameIndex = frameIndexStr.toInt(&convertedToIntSuccessfully);
if (!convertedToIntSuccessfully) {
qWarning() << "Failed to convert frame index" << frameIndexStr << "to an int";
return -1;
}
return frameIndex;
}
int SpriteImageProvider::parseFrameWidth(const QString &frameWidthStr) const
{
bool convertedToIntSuccessfully = false;
const int frameWidth = frameWidthStr.toInt(&convertedToIntSuccessfully);
if (!convertedToIntSuccessfully) {
qWarning() << "Failed to convert frame width" << frameWidthStr << "to an int";
return -1;
}
return frameWidth;
}
int SpriteImageProvider::parseFrameHeight(const QString &frameHeightStr) const
{
bool convertedToIntSuccessfully = false;
const int frameHeight = frameHeightStr.toInt(&convertedToIntSuccessfully);
if (!convertedToIntSuccessfully) {
qWarning() << "Failed to convert frame height" << frameHeightStr << "to an int";
return -1;
}
return frameHeight;
}
It also supports "alpha keys" - white and magenta pixels in the source image will be turned into transparent pixels.
It's registered with the QML engine like this:
mEngine->addImageProvider("sprite", new SpriteImageProvider);
and used in QML like this:
import QtQuick 2.5
Image {
source: qsTr('image://sprite/sprites/hugh.png,%1').arg(_sceneItemComponent ? _sceneItemComponent.facingDirection : 0)
}
where facingDirection is the frame index to use. There are also other arguments you can pass that are mentioned in the .cpp file.

compare images using systemC

I wrote in this forum asking for help to solve this problem that took ame a lot of my time,i write my first program using systemC, I will expain my aim as much as I can , I stored 2 matrix of pixel value of image in two different text files, I write a systemC code that load two matrix and apply somme of absolute difference, if number of different superior of a Threshold the code displays message (motion).
My code composed of two modules, the first module check if there a number stored in a text file, if yes this Module will automates the other module to load the two matrix and compare them, I really need this code for my project graduation any help or suggestion.
#include "systemC.h"
#include "string.h"
#include "stdio.h"
#include"stdlib.h"
#include <time.h>
#include <math.h> /* fabs */
#include <fstream>
#include <iostream>
#include <fstream>
using namespace std;
#define _CRT_SECURE_NO_WARNINGS
_CRT_SECURE_NO_WARNINGS
double elapsed;
int H = 0;
int D = 0;
int a, b;
int in = false;
int L = 0;
char *mode1 = "r";
char *mode2 = "w";
int i, j, k;
int rows1, cols1, rows2, cols2;
bool fileFound = false;
FILE *SwitchContext;
FILE *image1;
FILE *image2;
FILE *image3;
int sum = 0;
clock_t start = clock();
SC_MODULE(synchronization)
{
sc_in<bool>sig ;
SC_CTOR(synchronization)
{
SC_METHOD(synchroprocess)
}
void synchroprocess()
{
cout << "\n Running Automation";
SwitchContext = fopen("F:/SWITCH CONTEXT.txt", mode2);
fscanf(SwitchContext, "%d", &L);
while (L != 0)
{
cout << "waiting...";
}
sig == true;
}
};
SC_MODULE(imageProcess)
{
sc_in<bool>sig;
SC_CTOR(imageProcess)
{
SC_METHOD(MotionDetector)
sensitive(sig);
}
void MotionDetector()
{
image3 = fopen("F:/image3.txt", mode2);
do
{
char *mode1 = "r";
char *mode2 = "w";
image1 = fopen("F:/image1.txt", mode1);
if (!image1)
{
printf("File Not Found!!\n");
fileFound = true;
}
else
fileFound = false;
}
while (fileFound);
do
{
image2 = fopen("F:/image2.txt", mode1);
if (!image2)
{
printf("File Not Found!!\n");
fileFound = true;
}
else
fileFound = false;
}
while (fileFound);
rows1 = rows2 = 384;
cols1 = cols2 = 512;
int **mat1 = (int **)malloc(rows1 * sizeof(int*));
for (i = 0; i < rows1; i++)
mat1[i] = (int *)malloc(cols1 * sizeof(int));
i = 0;
int **mat2 = (int **)malloc(rows2 * sizeof(int*));
for (i = 0; i < rows2; i++)
mat2[i] = (int *)malloc(cols2 * sizeof(int));
i = 0;
while (!feof(image1))
{
for (i = 0; i < rows1; i++)
{
for (j = 0; j < cols1; j++)
fscanf(image1, "%d%", &mat1[i][j]);
}
}
i = 0;
j = 0;
while (!feof(image2))
{
for (i = 0; i < rows2; i++)
{
for (j = 0; j < cols2; j++)
fscanf(image2, "%d%", &mat2[i][j]);
}
}
i = 0;
j = 0;
printf("\n\n");
for (i = 0; i < rows1; i++)
{
for (j = 0; j < cols1; j++) {
a = abs(mat1[i][j] = mat2[i][j]);
b = b + a;
}
}
i = j = 0;
D = b / 196608;
if (D > 0.9)
{
printf("%d,&K");
printf("MOTION...DETECTED");
getchar();
sc_pause;
for (i = 0; i < rows1; i++) {
for (j = 0; j < cols1; j++)
{
fprintf(image3, "%d ", mat2[i][j]);
}
fprintf(image3, "\n");
}
printf("\n Image Saved....");
std::ofstream mon_fichier("F:\toto.txt");
mon_fichier << elapsed << '\n';
}
fclose(image1);
fclose(image2);
fclose(image3);
clock_t end = clock();
elapsed = ((double)end - start) / CLOCKS_PER_SEC;
printf("time is %f", elapsed);
}
};
int sc_main(int argc, char* argv[])
{
imageProcess master("EE2");
master.MotionDetector();
sc_start();
return(0);
}
What you did is basically wrong.
You copy pasted code to SC_MODULE, this code is simple C code
(Do not mix C and C++ files)
This is not how you use clock
What you should do:
You need to check if your algorithm works, for this you do not need SystemC at all
Then you can replace data types with HW one and check if it still works
Then you have to find which data interface is used in HW and how to use this interface
Then you have to tweak your alg. to work with this interface (There you can use SC_MODULE, sc ports etc...)
Also take look at SC_CTHREAD, you will need it.
Without any informations about target platform I can not provide any other help.

Teaching myself OOP in C++

So I've been working on this program for the last month. The original code is from this tutorial https://www.youtube.com/watch?v=KjHKwCZyAhQ&list=PLHm_I0tE5kKPPWXkTTtOn8fkcwEGZNETh&index=3
However, I thought I would turn it into an object oriented program before I went on. Doing rather than copying is the best way to learn. The code generated a bmp file before i divided it up, but not anymore. The program executes but it doesn't create a file. Additionally I added Hello World in my .cpp files to see if they were even being executed and it looks like they aren't. I realize in copying this that I have a lot of code, I think the problem is in the main file so hopefully if anyone is nice enough to help me they can pick it out much more quickly!
*edit
Also in the original code he had the strut as a global variable but wasn't sure which file to implement it in or even how to make something global in an OOP! Would I just put it in main above int main() ?
Output.h
#pragma once
#include "ProProcess.h" //this is just a bunch of preprocessor directives
//this program creats a single color bmp file using red, blue, and green (rgb)
class OutPut
{
public:
OutPut(const int height, std::string file_name, int dpi, int index);
~OutPut();
//savebmp_str(std::string* file_name, const int width, const int height, int dpi, int pixels, struct RGBtype);
//commented this out because I wasn't sure how I should pass all these values. Ultimately I used OutPut Object_Output in bmp.cpp so that these variables could be passed in there
const int Getwidth() { return width; }
const int Setwidth(const int x) { const int width = x; }
private:
struct RGBtype //Could be a global variable but I dont know which file to put it in
{
int r;
int g;
int b;
};
const int width = 1960; //window size
const int height = 1080;
int dpi = 72;
int number_of_pixels = width*height;
int index;
const char* file_name = "Scene.bmp";
RGBtype *pixels = new RGBtype[number_of_pixels];//creates an array so that each pixel is comprised of a mix of rgb
};
Output.cpp
#include "OutPut.h"
#include "ProProcess.h"
OutPut::OutPut(const int height, std::string file_name, int dpi, int index)
{
OutPut::RGBtype color;
for (int x = 0; x < height; x++) //nested for loop that draws out each pixel totalling 1920x1080 in all
{
for (int y = 0; y < width; y++)
{
index = y*height + x;
pixels[index].r = 311;//changing the number here changes the color
pixels[index].g = 311;
pixels[index].b = 311;
}
}
std::cout << "Hello World";
}
OutPut::~OutPut()
{
}
BMP.h
#pragma once
#include "ProProcess.h"
#include "OutPut.h"
struct RGBtype
{
int r;
int g;
int b;
};
class BMP
{
public:
BMP(const char *filename, int passed_width, int passed_height, int dpi, RGBtype* data);
~BMP();
private:
OutPut Object_Output(std::string* file_name, const int width, const int height, int dpi, int pixels, struct RGBtype);//this is to pass the variables declared in output.h so bmp.h and bmp.cpp can use them too. Not sure how I would even verify i am doing this properly!
//const char* savebmp_str();
int passed_width;
int passed_height;
int dpi;
RGBtype *data;
};
BMP.cpp
#include "BMP.h"
#include "ProProcess.h"
#include "OutPut.h"
BMP::BMP(const char *filename, int passed_width, int passed_height, int dpi, RGBtype *data)
{
std::cout << passed_height;
FILE *pFile;
int k = passed_width*passed_height;
std::cout << "The value k is" << k;
int s = 4 * k;
int filesize = 54 + s; //s is a function of width and height
double factor = 39.375;
int m = static_cast<int>(factor);
int ppm = dpi*m;
unsigned char bmpfileheader[14] = { 'B','M',0,0,0,0 ,0,0,0,0, 54,0,0,0 }; //B and M are case sensitive. They make a bmp file
unsigned char bmpinfoheader[40] = { 40,0,0,0, 0,0,0,0 ,0,0,0,0, 1,0,24,0 };// the header size 14 and 40 are part of the BMP format
bmpfileheader[2] = (unsigned char)(filesize);
bmpfileheader[3] = (unsigned char)(filesize >> 8);
bmpfileheader[4] = (unsigned char)(filesize >> 16);
bmpfileheader[5] = (unsigned char)(filesize >> 24);
bmpinfoheader[4] = (unsigned char)(passed_width);
bmpinfoheader[5] = (unsigned char)(passed_width >> 8);
bmpinfoheader[6] = (unsigned char)(passed_width >> 16);
bmpinfoheader[7] = (unsigned char)(passed_width >> 24);
bmpinfoheader[8] = (unsigned char)(passed_height);
bmpinfoheader[9] = (unsigned char)(passed_height >> 8);
bmpinfoheader[10] = (unsigned char)(passed_height >> 16);
bmpinfoheader[11] = (unsigned char)(passed_height >> 24);
bmpinfoheader[21] = (unsigned char)(s);
bmpinfoheader[22] = (unsigned char)(s >> 8);
bmpinfoheader[23] = (unsigned char)(s >> 16);
bmpinfoheader[24] = (unsigned char)(s >> 24);
bmpinfoheader[25] = (unsigned char)(ppm);
bmpinfoheader[26] = (unsigned char)(ppm >> 8);
bmpinfoheader[27] = (unsigned char)(ppm >> 16);
bmpinfoheader[28] = (unsigned char)(ppm >> 24);
bmpinfoheader[29] = (unsigned char)(ppm);
bmpinfoheader[30] = (unsigned char)(ppm >> 8);
bmpinfoheader[31] = (unsigned char)(ppm >> 16);
bmpinfoheader[32] = (unsigned char)(ppm >> 24);
pFile = fopen(filename, "wb");
fwrite(bmpfileheader, sizeof(char), 14, pFile);
fwrite(bmpinfoheader, sizeof(char), 40, pFile);
for (int i = 0; i < k; i++)
{
RGBtype rgb = data[i];
double red = (data[i].r);
double green = (data[i].g);
double blue = (data[i].b);
int color[3] = { (int)floor(blue), (int)floor(green), (int)floor(red) };
fwrite(color, 1, 3, pFile);
}
fclose(pFile);
std::cout << "Hello World";
}
BMP::~BMP()
{
}
main.cpp
#include <iostream>
#include "OutPut.h"
#include "ProProcess.h"
#include "BMP.h"
int main()
{
OutPut Pixel_gen();
BMP BMP_Format_Maker();
OutPut Object_Output();
system("Pause");
return 0;
}

OpenKinect acquire raw depth image

I am trying to use the example code from here.
I have made some changes in order to save the images to the computer. When I read the data in MATLAB it seems like values that should be 0 are set to 2047, and overall it does not seem to be correct when I reconstruct the 3D points using the default intrinsic camera parameters.
What I want to achieve is to save the images so that I can use
img = single(imread(depth.png'))/ 1000
and have the depth values in meters, and pixels with no measurements should be zero.
It is the Kinect V1 by the way.
Here is the code with comments where I have tried to change.
#include "libfreenect.hpp"
#include <iostream>
#include <vector>
#include <cmath>
#include <pthread.h>
#include <cv.h>
#include <cxcore.h>
#include <highgui.h>
using namespace cv;
using namespace std;
class myMutex {
public:
myMutex() {
pthread_mutex_init( &m_mutex, NULL );
}
void lock() {
pthread_mutex_lock( &m_mutex );
}
void unlock() {
pthread_mutex_unlock( &m_mutex );
}
private:
pthread_mutex_t m_mutex;
};
// Should one use FREENECT_DEPTH_REGISTERED instead of FREENECT_DEPTH_11BIT?
class MyFreenectDevice : public Freenect::FreenectDevice {
public:
MyFreenectDevice(freenect_context *_ctx, int _index)
: Freenect::FreenectDevice(_ctx, _index), m_buffer_depth(FREENECT_DEPTH_11BIT),
m_buffer_rgb(FREENECT_VIDEO_RGB), m_gamma(2048), m_new_rgb_frame(false),
m_new_depth_frame(false), depthMat(Size(640,480),CV_16UC1),
rgbMat(Size(640,480), CV_8UC3, Scalar(0)),
ownMat(Size(640,480),CV_8UC3,Scalar(0)) {
for( unsigned int i = 0 ; i < 2048 ; i++) {
float v = i/2048.0;
v = std::pow(v, 3)* 6;
m_gamma[i] = v*6*256;
}
}
// Do not call directly even in child
void VideoCallback(void* _rgb, uint32_t timestamp) {
std::cout << "RGB callback" << std::endl;
m_rgb_mutex.lock();
uint8_t* rgb = static_cast<uint8_t*>(_rgb);
rgbMat.data = rgb;
m_new_rgb_frame = true;
m_rgb_mutex.unlock();
};
// Do not call directly even in child
void DepthCallback(void* _depth, uint32_t timestamp) {
std::cout << "Depth callback" << std::endl;
m_depth_mutex.lock();
uint16_t* depth = static_cast<uint16_t*>(_depth);
// Here I use memcpy instead so I can use uint16
// memcpy(depthMat.data,depth,depthMat.rows*depthMat.cols*sizeof(uint16_t));
depthMat.data = (uchar*) depth;
m_new_depth_frame = true;
m_depth_mutex.unlock();
}
bool getVideo(Mat& output) {
m_rgb_mutex.lock();
if(m_new_rgb_frame) {
cv::cvtColor(rgbMat, output, CV_RGB2BGR);
m_new_rgb_frame = false;
m_rgb_mutex.unlock();
return true;
} else {
m_rgb_mutex.unlock();
return false;
}
}
bool getDepth(Mat& output) {
m_depth_mutex.lock();
if(m_new_depth_frame) {
depthMat.copyTo(output);
m_new_depth_frame = false;
m_depth_mutex.unlock();
return true;
} else {
m_depth_mutex.unlock();
return false;
}
}
private:
// Should it be uint16_t instead or even higher?
std::vector<uint8_t> m_buffer_depth;
std::vector<uint8_t> m_buffer_rgb;
std::vector<uint16_t> m_gamma;
Mat depthMat;
Mat rgbMat;
Mat ownMat;
myMutex m_rgb_mutex;
myMutex m_depth_mutex;
bool m_new_rgb_frame;
bool m_new_depth_frame;
};
int main(int argc, char **argv) {
bool die(false);
string filename("snapshot");
string suffix(".png");
int i_snap(0),iter(0);
Mat depthMat(Size(640,480),CV_16UC1);
Mat depthf (Size(640,480),CV_8UC1);
Mat rgbMat(Size(640,480),CV_8UC3,Scalar(0));
Mat ownMat(Size(640,480),CV_8UC3,Scalar(0));
// The next two lines must be changed as Freenect::Freenect
// isn't a template but the method createDevice:
// Freenect::Freenect<MyFreenectDevice> freenect;
// MyFreenectDevice& device = freenect.createDevice(0);
// by these two lines:
Freenect::Freenect freenect;
MyFreenectDevice& device = freenect.createDevice<MyFreenectDevice>(0);
namedWindow("rgb",CV_WINDOW_AUTOSIZE);
namedWindow("depth",CV_WINDOW_AUTOSIZE);
device.startVideo();
device.startDepth();
while (!die) {
device.getVideo(rgbMat);
device.getDepth(depthMat);
// Here I save the depth images
std::ostringstream file;
file << filename << i_snap << suffix;
cv::imwrite(file.str(),depthMat);
cv::imshow("rgb", rgbMat);
depthMat.convertTo(depthf, CV_8UC1, 255.0/2048.0);
cv::imshow("depth",depthf);
if(iter >= 1000) break;
iter++;
}
device.stopVideo();
device.stopDepth();
return 0;
}
Thanks in advance!
Erik
I dont have any experience with OpenKinect in particular; but should your depth buffer be uint16?
std::vector<uint8_t> m_buffer_depth;
Also; for Matlab, do check if the image that you are reading is a uint16 or uint8. If its the latter then convert it to uint16
uint16(imread('depth.png'));
Sorry couldn't help more. Hope this helps.
The values you have are the raw depth values. You need to remap those into MM for the numbers to make sense. Kinect 1 can see up to 10 meters. So I would go with raw_values/2407*10000.
If the values are saturated at 2047, you are probably using the FREENECT_DEPTH_11BIT_PACKED depth format.
For work in Matlab, it is always easier to use FREENECT_DEPTH_MM or FREENECT_DEPTH_REGISTERED.
Enjoy.