Read pixel of a L*a*b color space image

From opencv documentation i found that the Lab* color space has limited value for each variable as following :

   0 < L < 100
-127 < a < 127 
-127 < b < 127

I wrote a code which read and convert an image of type BGR to Lab* color space. When i show value of L,a and b i found that values are out of range (all of them)

For example in a pixel (y,x) the value of b is 150 but from the opencv 2.4.13 documentation b must be between -127 and 127. The code is the following :

   #include <opencv2/core/core.hpp>
   #include <opencv2/highgui/highgui.hpp>
   #include "opencv2/imgproc/imgproc.hpp"
   #include <iostream>

   using namespace cv;
   using namespace std;


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

    Mat input, Lab_img;

    input = imread("E:\\Walid\\Images\\b2.jpg");


    cvtColor(input, Lab_img, CV_BGR2Lab);
    namedWindow("ORIGINAL", WINDOW_AUTOSIZE);
    namedWindow("Lab", WINDOW_AUTOSIZE);



    for (int y = 0; y < Lab_img.rows; y++)
    {
        for (int x = 0; x < Lab_img.cols; x++)
        {
            Vec3b intensity = Lab_img.at<Vec3b>(y, x);
            double L = intensity.val[0];
            double a = intensity.val[1];
            double b = intensity.val[2];
            cout << b << std::endl;
        }
    }


    imshow("ORIGINAL", input);
    imshow("Lab", Lab_img);

    waitKey(0);
    return 0;
   }

Here is reference for cvtColor In section RGB <-> CIE L*a*b* it says:

This outputs 0 <= L <= 100, -127 <= a <= 127, -127 <= b <= 127 . The values are then converted to the destination data type: For 8-bit images L = L*255/100, a = a + 128, b = b + 128.