So long story short, I'm looking for an OpenCV model that does training on multiple image sets based on color, not on shape. Like a FaceRegonition algorithm (Fisherfacerecognizer, EigenFaces, Lbphfacerecognizer etc) but instead of looking for facials features, it looks for colors and then, using an input image, makes a prediction based on previous training.

From what I understood, SVM (Support Vector Machine) would be the preferable approach. However, I have no idea on how should I train it. I'll leave you with some code and a concrete case:

I'm doing a "Banana Ripeness Detector" using OpenCV & C++ where you load a picture of a banana and it tells you how "ripe" it is. For this purpose, I have a training_models folder with 3 subfolders "unripe_bananas", "ripe_bananas" and "overripe_bananas", each with multiple pictures of bananas.

Then I have a CSV in order to read all this pictures inside the program inside a vector called "images". The number after semicolon represents the label (unripe_bananas -> label 1, ripe_bananas -> label 2, overripe_bananas -> label 3)

resources/training_bananas/unripe_bananas/banana1.jpg;1
resources/training_bananas/unripe_bananas/banana2.jpg;1
resources/training_bananas/unripe_bananas/banana3.jpg;1
resources/training_bananas/unripe_bananas/banana4.jpg;1
resources/training_bananas/unripe_bananas/banana5.jpg;1
resources/training_bananas/unripe_bananas/banana6.jpg;1
resources/training_bananas/unripe_bananas/banana7.jpg;1
resources/training_bananas/ripe_bananas/banana8.jpg;2
resources/training_bananas/ripe_bananas/banana9.jpg;2
resources/training_bananas/ripe_bananas/banana10.jpg;2
resources/training_bananas/ripe_bananas/banana11.jpg;2
...

And this is the code. The banana detection part in a picture is already handled using "banana_detect_cascade.xml" (works similary to the haarcascade_frontalface_alt.xml cascade)

#include "opencv2/objdetect/objdetect.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/ml.hpp"
#include <iostream>
#include <stdio.h>
#include <fstream>

using namespace std;
using namespace cv;
using namespace ml;

void detectAndDisplayBanana(Mat frame);
/** Global variables */
String banana_cascade_name = "resources//banana_detect_cascade.xml";
CascadeClassifier banana_cascade;
string window_name = "Capture - Banana Ripe Detection";

// Reading the CSV containing banana pictures for training
// Numbers after semicolon (;) are labels

static void read_csv(const string filename, vector<Mat>& images, vector<int>& labels, char separator = ';')
{
    ifstream file(filename, ios::in);
    if (file)
    {
        string line, path, classLabel;
        while (getline(file, line))
        {
            stringstream liness(line);
            getline(liness, path, separator);
            getline(liness, classLabel);
            if (!path.empty() && !classLabel.empty())
            {
                images.push_back(imread(path, ImreadModes::IMREAD_COLOR));
                labels.push_back(atoi(classLabel.c_str()));
            }
        }
    }
}
int im_width;
int im_height;
Ptr<SVM> model;




int main(int argc, const char** argv)
{

    Mat frame;
    //-- 1. Load the cascades
    if (!banana_cascade.load(banana_cascade_name)) { printf("--(!)Error loading\n"); return -1; };
    string fileName = string("resources//csv.ext");

    
    vector<Mat> images; // the vector with the training images
    vector<int> labels; // the vector with the labels for the training images
    read_csv(fileName, images, labels);

    im_width = images[0].cols;
    im_height = images[0].rows;
    model = SVM::create();
    

    ///////// .... HOW TO TRAIN ?????


    //-- 3. Read the image

    while (true)
    {
        frame = imread("banana_example.jpg");
        //-- 3. Apply the classifier to the frame
        if (!frame.empty())
        {
            detectAndDisplayBanana(frame);
        }
        else
        {
            printf("No banana image!"); break;
        }
        int c = waitKey(10);
        if ((char)c == 'c') { break; }
    }

    return 0;
}


void detectAndDisplayBanana(Mat frame)
{
    // This functions detect bananas in a picture - already handled

    string ripeStage;
    std::vector<Rect> bananas;
    Mat frame_gray;
    cvtColor(frame, frame_gray, COLOR_BGR2GRAY);
    equalizeHist(frame_gray, frame_gray);
    //-- Detect bananas


    banana_cascade.detectMultiScale(frame_gray, bananas, 1.1, 2, 0 | 1, Size(30, 30));
    for (size_t i = 0; i < bananas.size(); i++)
    {
        Rect banana_i = bananas[i];
        Mat banana = frame_gray(banana_i);
        Mat banana_resized;
    cv:resize(banana, banana_resized, Size(im_width, im_height), 1.0, 1.0, InterpolationFlags::INTER_CUBIC);
        
        int ripeLabel = model->predict(banana_resized);

        switch (ripeLabel)
        {
        case 1:
            ripeStage = "Unripe";
            break;
        case 2:
            ripeStage = "Ripe";
            break;
        case 3:
            ripeStage = "Overripe";
            break;
        default:
            ripeStage = "Unknown";
        }



        
        
        string box_text = format("%s %f", ripeStage.c_str());
        cout << box_text << endl;
        
    }
    //-- Show what you got
    imshow(window_name, frame);
}