edges_sub_pix — Extract sub-pixel precise edges using Deriche, Lanser, Shen, or Canny filters.
edges_sub_pix detects step edges using recursively implemented filters (according to Deriche, Lanser and Shen) or the conventionally implemented “derivative of Gaussian” filter (using filter masks) proposed by Canny. Thus, the following edge operators are available:
'deriche1', 'lanser1', 'deriche2', 'lanser2', 'shen', 'mshen', 'canny', 'sobel', and 'sobel_fast'
The extracted edges are returned as sub-pixel precise XLD contours in Edges. For all edge operators except 'sobel_fast', the following attributes are defined for each edge point (see get_contour_attrib_xld):
|'angle'||Direction of the normal vectors to the contour (oriented such that the normal vectors point to the right side of the contour as the contour is traversed from start to end point; the angles are given with respect to the row axis of the image.)|
|'response'||Edge amplitude (gradient magnitude)|
The “filter width” (i.e., the amount of smoothing) can be chosen arbitrarily for all edge operators except 'sobel' and 'sobel_fast', and can be estimated by calling info_edges for concrete values of the parameter Alpha. For all filters (Deriche, Lanser and Shen filters), the “filter width” decreases for increasing Alpha. The only exception is the Canny filter, where an increasing Alpha also causes an increase of the “filter width”. “Wide” filters exhibit a larger invariance to noise, but also a decreased ability to detect small details. Non-recursive filters, such as the Canny filter, are realized using filter masks, and thus the execution time increases for increasing filter width. In contrast, the execution time for recursive filters does not depend on the filter width. Thus, arbitrary filter widths are possible using the Deriche, Lanser and Shen filters without increasing the run time of the operator. The resulting advantage in speed compared to the Canny operator naturally increases for larger filter widths. As border treatment, the recursive operators assume that the images to be zero outside of the image, while the Canny operator repeats the gray value at the image's border. The signal-noise-ratio of the filters is comparable for the following choices of Alpha:
Alpha('lanser1') = Alpha('deriche1'), Alpha('deriche2') = Alpha('deriche1') / 2, Alpha('lanser2') = Alpha('deriche2'), Alpha('shen') = Alpha('deriche1') / 2, Alpha('mshen') = Alpha('shen'), Alpha('canny') = 1.77 / Alpha('deriche1').The originally proposed recursive filters ('deriche1', 'deriche2', 'shen') return a biased estimate of the amplitude of diagonal edges. This bias is removed in the corresponding modified version of the operators ('lanser1', 'lanser2' and 'mshen'), while maintaining the same execution speed.
For relatively small filter widths (11 x 11), i.e., for Alpha ('lanser2' = 0.5), all filters yield similar results. Only for “wider” filters differences begin to appear: the Shen filters begin to yield qualitatively inferior results. However, they are the fastest of the implemented operators that supprt arbitrary mask sizes, closely followed by the Deriche operators. The two Sobel filters, which use a fixed mask size of (3 x 3), are faster than the other filters. Of these two, the filter 'sobel_fast' is significantly faster than 'sobel'.
edges_sub_pix links the edge points into edges by using an algorithm similar to a hysteresis threshold operation, which is also used in lines_gauss. Points with an amplitude larger than High are immediately accepted as belonging to an edge, while points with an amplitude smaller than Low are rejected. All other points are accepted as edges if they are connected to accepted edge points (see also lines_gauss and hysteresis_threshold).
Because edge extractors are often unable to extract certain junctions, a mode that tries to extract these missing junctions by different means can be selected by appending '_junctions' to the values of Filter that are described above. This mode is analogous to the mode for completing junctions that is available in lines_gauss.
The edge operator 'sobel_fast' has the same semantics as all the other edge operators. Internally, however, it is based on significantly simplified variants of the individual processing steps (hysteresis thresholding, edge point linking, and extraction of the subpixel edge positions). Therefore, 'sobel_fast' in some cases may return slightly less accurate edge positions and may select different edge parts.
edges_sub_pix can be executed on OpenCL devices for the filter types 'canny' and 'sobel_fast'. This will require up to width*height*29 bytes of pinned memory. Since allocating memory is an expensive operation, it would make sense to set the pinned memory cache to at least this size (using set_compute_device_param for parameter 'pinned_mem_cache_capacity', or to disable pinned memory completely (using set_compute_device_param for parameter 'alloc_pinned'), in which case the normal memory cache is used. Note that the results can vary from the CPU implementation.
Since edges_sub_pix uses Gauss convolution internally for the 'canny' filter, the same limitations for OpenCL apply as for derivate_gauss: Alpha must be chosen small enough that the required filter mask is less than 129 pixels in size. Also, edges_sub_pix is not available on OpenCL devices for HALCON XL, as double precision floating point arithmetic would be required, but OpenCL devices are optimized for single precision arithmetic.
Note that filter operators may return unexpected results if an image with a reduced domain is used as input. Please refer to the chapter Filters.
Edge operator to be applied.
Default value: 'canny'
List of values: 'canny', 'canny_junctions', 'deriche1', 'deriche1_junctions', 'deriche2', 'deriche2_junctions', 'lanser1', 'lanser1_junctions', 'lanser2', 'lanser2_junctions', 'mshen', 'mshen_junctions', 'shen', 'shen_junctions', 'sobel', 'sobel_fast', 'sobel_junctions'
List of values (for compute devices): 'canny', 'sobel_fast'
Filter parameter: small values result in strong smoothing, and thus less detail (opposite for 'canny').
Default value: 1.0
Suggested values: 0.1, 0.2, 0.3, 0.4, 0.5, 0.7, 0.9, 1.1
Typical range of values: 0.2 ≤ Alpha ≤ 50.0
Minimum increment: 0.01
Recommended increment: 0.1
Restriction: Alpha > 0.0
Lower threshold for the hysteresis threshold operation.
Default value: 20
Suggested values: 5, 10, 15, 20, 25, 30, 40
Typical range of values: 1 ≤ Low ≤ 255
Minimum increment: 1
Recommended increment: 5
Restriction: Low > 0
Upper threshold for the hysteresis threshold operation.
Default value: 40
Suggested values: 10, 15, 20, 25, 30, 40, 50, 60, 70
Typical range of values: 1 ≤ High ≤ 255
Minimum increment: 1
Recommended increment: 5
Restriction: High > 0 && High >= Low
Let A be the number of pixels in the domain of Image. Then the runtime complexity is O(A*Alpha) for the Canny filter and O(A) for the recursive Lanser, Deriche, and Shen filters.
The amount of temporary memory required is dependent on the height H of the domain of Image and the width W of Image. Let S = W*H, then edges_sub_pix requires at least 60*S bytes of temporary memory during execution for all edge operators except 'sobel_fast'. For 'sobel_fast', at least 9*S bytes of temporary memory are required.
edges_sub_pix returns 2 (H_MSG_TRUE) if all parameters are correct and no error occurs during execution. If the input is empty the behavior can be set via set_system('no_object_result',<Result>). If necessary, an exception is raised.
segment_contours_xld, gen_polygons_xld, select_shape_xld
sobel_dir, frei_dir, kirsch_dir, prewitt_dir, robinson_dir, edges_image
info_edges, hysteresis_threshold, bandpass_image, lines_gauss, lines_facet
S.Lanser, W.Eckstein: “Eine Modifikation des Deriche-Verfahrens zur
Kantendetektion”; 13. DAGM-Symposium, München; Informatik
Fachberichte 290; Seite 151 - 158; Springer-Verlag; 1991.
S.Lanser: “Detektion von Stufenkanten mittels rekursiver Filter nach Deriche”; Diplomarbeit; Technische Universität München, Institut für Informatik, Lehrstuhl Prof. Radig; 1991.
J.Canny: “Finding Edges and Lines in Images”; Report, AI-TR-720; M.I.T. Artificial Intelligence Lab., Cambridge; 1983.
J.Canny: “A Computational Approach to Edge Detection”; IEEE Transactions on Pattern Analysis and Machine Intelligence; PAMI-8, vol. 6; S. 679-698; 1986.
R.Deriche: “Using Canny's Criteria to Derive a Recursively Implemented Optimal Edge Detector”; International Journal of Computer Vision; vol. 1, no. 2; S. 167-187; 1987.
R.Deriche: “Optimal Edge Detection Using Recursive Filtering”; Proc. of the First International Conference on Computer Vision, London; S. 501-505; 1987.
R.Deriche: “Fast Algorithms for Low-Level Vision”; IEEE Transactions on Pattern Analysis and Machine Intelligence; PAMI-12, no. 1; S. 78-87; 1990.
S.Castan, J.Zhao und J.Shen: “Optimal Filter for Edge Detection Methods and Results”; Proc. of the First European Conference on Computer Vision, Antibes; Lecture Notes on computer Science; no. 427; S. 12-17; Springer-Verlag; 1990.