ó  ‰\c@sGddlZddlmZmZmZmZddlmZm Z m Z m Z ddl m Z ddlmZddlmZejdƒZd d d d d d d d d d ddddddgZxbeddƒD]QZxHeeeeƒ eeeƒdƒD]Zded ed ef Ip + threshold``. Also stands for the n in ``FAST-n`` corner detector. fast_threshold : float, optional The ``threshold`` parameter in ``feature.corner_fast``. Threshold used to decide whether the pixels on the circle are brighter, darker or similar w.r.t. the test pixel. Decrease the threshold when more corners are desired and vice-versa. harris_k : float, optional The `k` parameter in `skimage.feature.corner_harris`. Sensitivity factor to separate corners from edges, typically in range ``[0, 0.2]``. Small values of `k` result in detection of sharp corners. downscale : float, optional Downscale factor for the image pyramid. Default value 1.2 is chosen so that there are more dense scales which enable robust scale invariance for a subsequent feature description. n_scales : int, optional Maximum number of scales from the bottom of the image pyramid to extract the features from. Attributes ---------- keypoints : (N, 2) array Keypoint coordinates as ``(row, col)``. scales : (N, ) array Corresponding scales. orientations : (N, ) array Corresponding orientations in radians. responses : (N, ) array Corresponding Harris corner responses. descriptors : (Q, `descriptor_size`) array of dtype bool 2D array of binary descriptors of size `descriptor_size` for Q keypoints after filtering out border keypoints with value at an index ``(i, j)`` either being ``True`` or ``False`` representing the outcome of the intensity comparison for i-th keypoint on j-th decision pixel-pair. It is ``Q == np.sum(mask)``. References ---------- .. [1] Ethan Rublee, Vincent Rabaud, Kurt Konolige and Gary Bradski "ORB: An efficient alternative to SIFT and SURF" http://www.vision.cs.chubu.ac.jp/CV-R/pdf/Rublee_iccv2011.pdf Examples -------- >>> from skimage.feature import ORB, match_descriptors >>> img1 = np.zeros((100, 100)) >>> img2 = np.zeros_like(img1) >>> np.random.seed(1) >>> square = np.random.rand(20, 20) >>> img1[40:60, 40:60] = square >>> img2[53:73, 53:73] = square >>> detector_extractor1 = ORB(n_keypoints=5) >>> detector_extractor2 = ORB(n_keypoints=5) >>> detector_extractor1.detect_and_extract(img1) >>> detector_extractor2.detect_and_extract(img2) >>> matches = match_descriptors(detector_extractor1.descriptors, ... detector_extractor2.descriptors) >>> matches array([[0, 0], [1, 1], [2, 2], [3, 3], [4, 4]]) >>> detector_extractor1.keypoints[matches[:, 0]] array([[ 42., 40.], [ 47., 58.], [ 44., 40.], [ 59., 42.], [ 45., 44.]]) >>> detector_extractor2.keypoints[matches[:, 1]] array([[ 55., 53.], [ 60., 71.], [ 57., 53.], [ 72., 55.], [ 58., 57.]]) g333333ó?iiôi g{®Gáz´?g{®Gáz¤?cCsg||_||_||_||_||_||_d|_d|_d|_ d|_ d|_ dS(N( t downscaletn_scalest n_keypointstfast_ntfast_thresholdtharris_ktNonet keypointstscalest responsest orientationst descriptors(tselfR R RRRR((s2lib/python2.7/site-packages/skimage/feature/orb.pyt__init__us          cCs2t|ƒ}tt||jd|jdtƒƒS(Nit multichannel(RtlistRR R tFalse(Rtimage((s2lib/python2.7/site-packages/skimage/feature/orb.pyt_build_pyramid…s cCs t||j|jƒ}t|ddƒ}t|ƒdkrtjd dtjƒtjd dtjƒtjd dtjƒfSt|j |ddƒ}||}t ||t ƒ}t |dd d |j ƒ}||dd…df|dd…dff}|||fS( Nt min_distanceiiitdtypetdistanceitmethodtk(ii(i(i(RRRRtlentnptzerostdoubleRtshapeRt OFAST_MASKRR(Rt octave_imaget fast_responseRtmaskRtharris_responseR((s2lib/python2.7/site-packages/skimage/feature/orb.pyt_detect_octaveŠs        0cCs°t|dƒ|j|ƒ}g}g}g}g}x¨tt|ƒƒD]”}tj||ƒ}|j|ƒ\} } } |j| |j|ƒ|j| ƒ|j|j|tj | j ddtj ƒƒ|j| ƒqGWtj |ƒ} tj |ƒ} tj |ƒ} tj |ƒ} | j d|jkrX| |_| |_| |_| |_nT| jƒddd…|j } | | |_| | |_| | |_| | |_dS(s¥Detect oriented FAST keypoints along with the corresponding scale. Parameters ---------- image : 2D array Input image. iiR Niÿÿÿÿ(R RtrangeR$R%tascontiguousarrayR.tappendR tonesR(tintptvstackthstackRRRRRtargsort(RRtpyramidtkeypoints_listtorientations_listt scales_listtresponses_listtoctaveR*RRRRt best_indices((s2lib/python2.7/site-packages/skimage/feature/orb.pytdetect¢s8  !        cCs„t|j|ddƒ}tj||dtjdddtƒ}tj||dtjdddtƒ}t|||ƒ}||fS(NR!iR tordertCtcopy(RR(R%tarrayR3RR'R (RR*RRR,R((s2lib/python2.7/site-packages/skimage/feature/orb.pyt_extract_octaveÓs   cCs5t|dƒ|j|ƒ}g}g}tj|ƒtj|jƒjtjƒ}x«tt|ƒƒD]—} || k} tj | ƒdkrftj || ƒ} || } | |j| :} || } |j | | | ƒ\}}|j |ƒ|j |ƒqfqfWtj |ƒjtjƒ|_tj|ƒ|_dS(sExtract rBRIEF binary descriptors for given keypoints in image. Note that the keypoints must be extracted using the same `downscale` and `n_scales` parameters. Additionally, if you want to extract both keypoints and descriptors you should use the faster `detect_and_extract`. Parameters ---------- image : 2D array Input image. keypoints : (N, 2) array Keypoint coordinates as ``(row, col)``. scales : (N, ) array Corresponding scales. orientations : (N, ) array Corresponding orientations in radians. iiN(R RR%tlogR tastypeR3R/R$tsumR0RCR1R4tviewtboolRR5tmask_(RRRRRR7tdescriptors_listt mask_listtoctavesR<t octave_maskR*toctave_keypointstoctave_orientationsRR,((s2lib/python2.7/site-packages/skimage/feature/orb.pytextractßs& +     cCst|dƒ|j|ƒ}g}g}g}g}g}x-tt|ƒƒD]}tj||ƒ} |j| ƒ\} } } t| ƒdkrÏ|j| ƒ|j| ƒ|jtjddtj ƒƒqMn|j | | | ƒ\} }|j| ||j |ƒ|j| |ƒ|j| |ƒ|j|j |tj | j ddtjƒƒ|j| ƒqMWt|ƒdkr‹tdƒ‚ntj|ƒ} tj|ƒ} tj|ƒ}tj|ƒ} tj|ƒjtj ƒ} | j d|jkr(| |_||_| |_| |_| |_na| jƒddd…|j }| ||_|||_| ||_| ||_| ||_dS( sûDetect oriented FAST keypoints and extract rBRIEF descriptors. Note that this is faster than first calling `detect` and then `extract`. Parameters ---------- image : 2D array Input image. iiiR snORB found no features. Try passing in an image containing greater intensity contrasts between adjacent pixels.Niÿÿÿÿ(ii(R RR/R$R%R0R.R1R&RHRCR R2R(R3t RuntimeErrorR4R5RGRRRRRRR6(RRR7R8R;R:R9RJR<R*RRRRR,RR=((s2lib/python2.7/site-packages/skimage/feature/orb.pytdetect_and_extractsV    !           ( t__name__t __module__t__doc__RRR.R>RCRPRR(((s2lib/python2.7/site-packages/skimage/feature/orb.pyR s]     1 6(ii(tnumpyR%t feature.utilRRRRtfeatureRRRRt transformRt _shared.utilsR torb_cyR R&R)t OFAST_UMAXR/titabstjR (((s2lib/python2.7/site-packages/skimage/feature/orb.pyts ""6/