连接组件算法输出

时间:2015-10-08 07:41:42

标签: c++ opencv image-processing

我目前正在实施一种识别图像中连接组件的算法。在执行自适应阈值处理后,我创建了另一个名为sub的Mat变量,我在其中更新其值,使得包含它们的值允许我使用更新' newObject'来识别哪些像素连接到哪个像素。变量。

当我完成这个时,我的预期输出是一个相对连接的图片,具有各种强度,可以看到哪些部分连接到哪个部分。然而,我的问题是我的输出是子变量只显示黑色图像。

我通过输入print语句检查当前值来检查值是否实际发生了变化。我还在if语句中将各种打印语句放在一点,并了解到它们确实进入了我更改子变量中像素值的条件。

我的问题是当我在子变量上执行imshow函数时,为什么输出图像是黑色图像。对不起,如果我在某处犯了错误,我相对较新。

for(int i = 0; i < detected_edges.rows; i++){
        for(int j = 0; j < detected_edges.cols; j++){
            if(detected_edges.at<uchar>(i,j) > 0){
                Point current = (i,j);
                if(i-1 < 0 && j-1 < 0){
                    newObject = newObject + 15;     //no values
                    sub.at<uchar>(current) = newObject;
                    printf("This is the value at this point: %u", sub.at<uchar>(current));
                }else if(i-1 >= 0 && j-1 < 0){
                    Point up = (i-1,j);
                    if(sub.at<uchar>(up) != 0){
                        sub.at<uchar>(current) = sub.at<uchar>(up);                 //only up has value
                    }else{
                        newObject = newObject + 15;                                             //no values
                        sub.at<uchar>(current) = newObject;
                    }
                }else if(i-1 < 0 && j-1 >= 0){
                    Point left = (i,j-1);
                    if(sub.at<uchar>(left) != 0){
                        sub.at<uchar>(current) = sub.at<uchar>(left);               //only left has value
                    }else{
                        newObject = newObject + 15;                                 //no values
                        sub.at<uchar>(current) = newObject;
                    }
                }else{
                    Point up = (i-1,j);
                    Point left = (i,j-1);
                    if(sub.at<uchar>(up) == 0 && sub.at<uchar>(left) == 0){

                        newObject = newObject + 15;                                             //no values
                        sub.at<uchar>(current) = newObject;
                    }else if(sub.at<uchar>(up) == sub.at<uchar>(left)){             //same value
                        sub.at<uchar>(current) = sub.at<uchar>(up);
                    }else if(sub.at<uchar>(up) != sub.at<uchar>(left) && (sub.at<uchar>(up) != 0 && sub.at<uchar>(left) != 0)){
                        sub.at<uchar>(current) = sub.at<uchar>(up);                 //different values follow upper's value
                    }else if((sub.at<uchar>(up) != 0 && sub.at<uchar>(left) == 0)){
                        sub.at<uchar>(current) = sub.at<uchar>(up);                 //only up has value
                    }else if(sub.at<uchar>(up) == 0 && sub.at<uchar>(left) != 0 ){
                        sub.at<uchar>(current) = sub.at<uchar>(left);               //only left has value
                    }
                }
            }
        }
    }

更新:我注意到只有我的第一行&#39; sub&#39; Mat变量由于某种原因正在改变值,即使循环上升到图像的最后一行。

1 个答案:

答案 0 :(得分:0)

图像是黑色的原因是因为我通过制作更多的对象来浏览灰度图像(0-255)的值范围。我使用2D数组执行了不同的实现,并且已经实现了算法。