public void AdaptiveThreshold()
{
const string imagePath = @"Files\PMC5055614_00002.jpg";
OpenCvImageProcesser _imageProcessing = new OpenCvImageProcesser();
(var img, var threshold) = _imageProcessing.AdaptiveThreshold(imagePath, false);
Assert.Equal(new int[] { 794, 596, 3 }, Shape(img));
Assert.Equal(new int[] { 794, 596 }, Shape(threshold));
using (var th = threshold.ToBitmap())
{
th.Save(@"Files\Output\PMC5055614_00002_threshold.png");
}
(var img_bg, var threshold_bg) = _imageProcessing.AdaptiveThreshold(imagePath, true);
Assert.Equal(new int[] { 794, 596, 3 }, Shape(img_bg));
Assert.Equal(new int[] { 794, 596 }, Shape(threshold_bg));
using (var th = threshold_bg.ToBitmap())
{
th.Save(@"Files\Output\PMC5055614_00002_threshold_bg.png");
}
img.Dispose();
threshold.Dispose();
img_bg.Dispose();
threshold_bg.Dispose();
}
public void FindContours()
{
const string imagePath = @"Files\PMC5055614_00002.jpg";
OpenCvImageProcesser _imageProcessing = new OpenCvImageProcesser();
(var img, var threshold) = _imageProcessing.AdaptiveThreshold(imagePath, false);
(var vertical_mask, var _) = _imageProcessing.FindLines(threshold, direction: "vertical");
(var horizontal_mask, var _) = _imageProcessing.FindLines(threshold); //, direction: "horizontal");
var contours = _imageProcessing.FindContours(vertical_mask, horizontal_mask);
var expected_contours = new List <(int, int, int, int)>()
{
(120, 242, 330, 2),
(120, 179, 327, 2),
(121, 305, 103, 2),
(268, 249, 83, 2),
(414, 376, 62, 2),
(307, 309, 62, 2),
(265, 376, 61, 2),
(340, 375, 62, 1),
(223, 309, 62, 1),
(121, 118, 326, 1)
};
Assert.Equal(expected_contours.Count, contours.Count);
Assert.Equal(expected_contours, contours);
(var img_bg, var threshold_bg) = _imageProcessing.AdaptiveThreshold(imagePath, true);
(var vertical_mask_bg, var _) = _imageProcessing.FindLines(threshold_bg, direction: "vertical");
(var horizontal_mask_bg, var _) = _imageProcessing.FindLines(threshold_bg, direction: "horizontal");
var contours_bg = _imageProcessing.FindContours(vertical_mask_bg, horizontal_mask_bg);
var expected_contours_bg = new List <(int, int, int, int)>()
{
(53, 415, 237, 319),
(306, 540, 235, 132),
(308, 416, 236, 122),
(52, 396, 491, 6),
(52, 407, 476, 5),
(242, 260, 214, 57),
(306, 705, 236, 5),
(307, 717, 236, 7),
(246, 181, 204, 5),
(122, 86, 191, 6)
};
Assert.Equal(expected_contours_bg.Count, contours_bg.Count);
Assert.Equal(expected_contours_bg, contours_bg);
img.Dispose();
threshold.Dispose();
vertical_mask.Dispose();
horizontal_mask.Dispose();
img_bg.Dispose();
threshold_bg.Dispose();
vertical_mask_bg.Dispose();
horizontal_mask_bg.Dispose();
}
public void FindLines()
{
const string imagePath = @"Files\PMC5055614_00002.jpg";
OpenCvImageProcesser _imageProcessing = new OpenCvImageProcesser();
// horizontal
(var img, var threshold) = _imageProcessing.AdaptiveThreshold(imagePath, false);
// find_lines: no region, horizontal
(var dmask, var lines) = _imageProcessing.FindLines(threshold);
var expected_lines = new List <(int, int, int, int)>()
{
(414, 377, 476, 377),
(265, 377, 326, 377),
(340, 376, 402, 376),
(191, 376, 252, 376),
(120, 368, 192, 368),
(120, 335, 192, 335),
(414, 329, 476, 329),
(265, 329, 326, 329),
(340, 328, 402, 328),
(191, 328, 252, 328),
(389, 310, 451, 310),
(307, 310, 369, 310),
(223, 310, 285, 310),
(121, 306, 224, 306),
(121, 273, 224, 273),
(389, 262, 451, 262),
(307, 262, 369, 262),
(223, 262, 285, 262),
(268, 259, 417, 259),
(267, 257, 417, 257),
(268, 250, 351, 250),
(120, 242, 450, 242),
(120, 210, 225, 210),
(389, 194, 450, 194),
(309, 194, 371, 194),
(224, 194, 285, 194),
(120, 180, 447, 180),
(245, 161, 312, 161),
(371, 147, 447, 147),
(120, 147, 312, 147),
(311, 132, 372, 132),
(121, 119, 447, 119),
(374, 85, 447, 85),
(121, 85, 314, 85),
(313, 71, 375, 71)
};
Assert.Equal(expected_lines.Count, lines.Count);
Assert.Equal(expected_lines, lines);
// find_lines: region, horizontal
(var dmask1, var lines1) = _imageProcessing.FindLines(threshold, new List <(int, int, int, int)>()
{
(0, 0, 200, 200)
});
var expected_lines1 = new List <(int, int, int, int)>()
{
(120, 180, 200, 180),
(120, 147, 200, 147),
(121, 118, 200, 118),
(121, 85, 200, 85)
};
Assert.Equal(expected_lines1.Count, lines1.Count);
Assert.Equal(expected_lines1, lines1);
img.Dispose();
threshold.Dispose();
dmask.Dispose();
dmask1.Dispose();
// vertical
(var img_bg, var threshold_bg) = _imageProcessing.AdaptiveThreshold(imagePath, true);
// find_lines: no region, vertical
(var dmask_v, var lines_v) = _imageProcessing.FindLines(threshold_bg, direction: "vertical");
var expected_lines2 = new List <(int, int, int, int)>()
{
(55, 734, 55, 659),
(195, 665, 195, 608),
(505, 661, 505, 606),
(393, 651, 393, 583),
(351, 627, 351, 573),
(308, 672, 308, 551),
(541, 659, 541, 540),
(55, 648, 55, 538),
(286, 601, 286, 527),
(185, 554, 185, 498),
(540, 525, 540, 441),
(287, 514, 287, 431),
(243, 481, 243, 425),
(308, 538, 308, 417),
(54, 527, 54, 416),
(478, 383, 478, 325),
(453, 317, 453, 258),
(452, 249, 452, 190)
};
Assert.Equal(expected_lines2.Count, lines_v.Count);
Assert.Equal(expected_lines2, lines_v);
// find_lines: region, vertical
(var dmask_v1, var lines_v1) = _imageProcessing.FindLines(threshold_bg, regions: new List <(int, int, int, int)>()
{
(0, 400, 190, 100)
}, direction: "vertical");
var expected_lines3 = new List <(int, int, int, int)>()
{
(54, 501, 54, 416)
};
Assert.Equal(expected_lines3.Count, lines_v1.Count);
Assert.Equal(expected_lines3, lines_v1);
(var dmask_v2, var lines_v2) = _imageProcessing.FindLines(threshold_bg, regions: new List <(int, int, int, int)>()
{
(0, 680, 190, 100)
}, direction: "vertical");
var expected_lines4 = new List <(int, int, int, int)>()
{
(55, 734, 55, 681)
};
Assert.Equal(expected_lines4.Count, lines_v2.Count);
Assert.Equal(expected_lines4, lines_v2);
img_bg.Dispose();
threshold_bg.Dispose();
dmask_v.Dispose();
dmask_v1.Dispose();
dmask_v2.Dispose();
}
public void FindJoints()
{
const string imagePath = @"Files\PMC5055614_00002.jpg";
OpenCvImageProcesser _imageProcessing = new OpenCvImageProcesser();
(var img, var threshold) = _imageProcessing.AdaptiveThreshold(imagePath, true);
(var vertical_mask, var _) = _imageProcessing.FindLines(threshold, direction: "vertical");
(var horizontal_mask, var _) = _imageProcessing.FindLines(threshold, direction: "horizontal");
var contours = _imageProcessing.FindContours(vertical_mask, horizontal_mask);
var table_bbox = _imageProcessing.FindJoints(contours, vertical_mask, horizontal_mask);
img.Dispose();
threshold.Dispose();
vertical_mask.Dispose();
horizontal_mask.Dispose();
var expected = new Dictionary <(float x1, float y1, float x2, float y2), List <(float, float)> >()
{
{
(53, 734, 290, 415),
new List <(float, float)>()
{
(55, 732),
(55, 721),
(55, 708),
(55, 695),
(55, 683),
(55, 671),
(55, 668),
(55, 660),
(195, 658),
(54, 646),
(195, 647),
(195, 634),
(55, 634),
(195, 621),
(55, 621),
(195, 610),
(55, 609),
(286, 598),
(55, 597),
(286, 585),
(55, 585),
(286, 573),
(54, 573),
(286, 561),
(54, 560),
(286, 549),
(54, 549),
(185, 549),
(54, 538),
(286, 536),
(185, 536),
(185, 524),
(54, 524),
(287, 512),
(185, 512),
(54, 512),
(185, 504),
(287, 501),
(185, 500),
(54, 500),
(287, 488),
(54, 487),
(287, 476),
(243, 476),
(54, 476),
(287, 465),
(243, 463),
(54, 463),
(54, 452),
(287, 452),
(243, 452),
(243, 443),
(287, 440),
(243, 440),
(54, 440),
(243, 427),
(54, 427),
(54, 416)
}
},
{
(306, 672, 541, 540),
new List <(float, float)>()
{
(309, 670),
(309, 658),
(540, 656),
(505, 657),
(393, 649),
(540, 645),
(505, 645),
(393, 645),
(308, 645),
(308, 633),
(540, 632),
(505, 632),
(393, 632),
(505, 624),
(540, 620),
(505, 620),
(393, 620),
(351, 621),
(308, 620),
(505, 612),
(308, 609),
(540, 608),
(505, 608),
(393, 608),
(351, 608),
(540, 597),
(351, 598),
(393, 597),
(308, 596),
(540, 585),
(393, 586),
(351, 584),
(308, 584),
(351, 576),
(308, 576),
(351, 574),
(540, 574),
(308, 573),
(540, 562),
(308, 561),
(540, 549)
}
},
