b0bdcbfdabcae1933980925c8f2438cbb2359a49,imgaug/augmenters/geometric.py,ElasticTransformation,_augment_keypoints,#ElasticTransformation#Any#Any#Any#Any#,3311

Before Change

        nb_images = len(keypoints_on_images)
        rss, alphas, sigmas, _orders, _cvals, _modes = self._draw_samples(
            nb_images, random_state)
        for i in sm.xrange(nb_images):
            kpsoi = keypoints_on_images[i]
            if not kpsoi.keypoints:
                // ElasticTransformation does not change the shape, hence we can
                // skip the below steps
                continue
            h, w = kpsoi.shape[0:2]
            dx, dy = self.generate_shift_maps(
                kpsoi.shape[0:2],
                alpha=alphas[i],
                sigma=sigmas[i],
                random_state=rss[i]
            )

            kps_aug = []
            for kp in kpsoi.keypoints:
                // dont augment keypoints if alpha/sigma are too low or if the
                // keypoint is outside of the image plane
                params_above_thresh = (
                        alphas[i] > self.KEYPOINT_AUG_ALPHA_THRESH
                        and sigmas[i] > self.KEYPOINT_AUG_SIGMA_THRESH)
                within_image_plane = (0 <= kp.x < w and 0 <= kp.y < h)
                if not params_above_thresh or not within_image_plane:
                    kps_aug.append(kp)
                else:
                    kp_neighborhood = kp.generate_similar_points_manhattan(
                        self.NB_NEIGHBOURING_KEYPOINTS,
                        self.NEIGHBOURING_KEYPOINTS_DISTANCE,
                        return_array=True
                    )

                    // We can clip here, because we made sure above that the
                    // keypoint is inside the image plane. Keypoints at the
                    // bottom row or right columns might be rounded outside
                    // the image plane, which we prevent here. We reduce
                    // neighbours to only those within the image plane as only
                    // for such points we know where to move them.
                    xx = np.round(kp_neighborhood[:, 0]).astype(np.int32)
                    yy = np.round(kp_neighborhood[:, 1]).astype(np.int32)
                    inside_image_mask = np.logical_and(
                        np.logical_and(0 <= xx, xx < w),
                        np.logical_and(0 <= yy, yy < h)
                    )
                    xx = xx[inside_image_mask]
                    yy = yy[inside_image_mask]

                    xxyy = np.concatenate(
                        [xx[:, np.newaxis], yy[:, np.newaxis]],
                        axis=1)

                    xxyy_aug = np.copy(xxyy).astype(np.float32)
                    xxyy_aug[:, 0] += dx[yy, xx]
                    xxyy_aug[:, 1] += dy[yy, xx]

                    med = ia.compute_geometric_median(xxyy_aug)
                    // uncomment to use average instead of median
                    // med = np.average(xxyy_aug, 0)
                    kps_aug.append(kp.deepcopy(x=med[0], y=med[1]))

            result[i] = kpsoi.deepcopy(keypoints=kps_aug)

        return result

    def _augment_polygons(self, polygons_on_images, random_state, parents,
                          hooks):

After Change

        result = keypoints_on_images
        nb_images = len(keypoints_on_images)
        samples = self._draw_samples(nb_images, random_state)
        gen = enumerate(zip(keypoints_on_images, samples.alphas, samples.sigmas,
                            samples.orders, samples.random_states))
        for i, (kpsoi, alpha, sigma, order, random_state_i) in gen:
            if not kpsoi.keypoints:
                // ElasticTransformation does not change the shape, hence we can
                // skip the below steps
                continue
            h, w = kpsoi.shape[0:2]
            dx, dy = self._generate_shift_maps(
                kpsoi.shape[0:2],
                alpha=alpha,
                sigma=sigma,
                random_state=random_state_i
            )

            kps_aug = []
            for kp in kpsoi.keypoints:
                // dont augment keypoints if alpha/sigma are too low or if the
                // keypoint is outside of the image plane
                params_above_thresh = (
                    alpha > self.KEYPOINT_AUG_ALPHA_THRESH
                    and sigma > self.KEYPOINT_AUG_SIGMA_THRESH)
                within_image_plane = (0 <= kp.x < w and 0 <= kp.y < h)
                if not params_above_thresh or not within_image_plane:
                    kps_aug.append(kp)
                else:
                    kp_neighborhood = kp.generate_similar_points_manhattan(
                        self.NB_NEIGHBOURING_KEYPOINTS,
                        self.NEIGHBOURING_KEYPOINTS_DISTANCE,
                        return_array=True
                    )

                    // We can clip here, because we made sure above that the
                    // keypoint is inside the image plane. Keypoints at the
                    // bottom row or right columns might be rounded outside
                    // the image plane, which we prevent here. We reduce
                    // neighbours to only those within the image plane as only
                    // for such points we know where to move them.
                    xx = np.round(kp_neighborhood[:, 0]).astype(np.int32)
                    yy = np.round(kp_neighborhood[:, 1]).astype(np.int32)
                    inside_image_mask = np.logical_and(
                        np.logical_and(0 <= xx, xx < w),
                        np.logical_and(0 <= yy, yy < h)
                    )
                    xx = xx[inside_image_mask]
                    yy = yy[inside_image_mask]

                    xxyy = np.concatenate(
                        [xx[:, np.newaxis], yy[:, np.newaxis]],
                        axis=1)

                    xxyy_aug = np.copy(xxyy).astype(np.float32)
                    xxyy_aug[:, 0] += dx[yy, xx]
                    xxyy_aug[:, 1] += dy[yy, xx]

                    med = ia.compute_geometric_median(xxyy_aug)
                    // uncomment to use average instead of median
                    // med = np.average(xxyy_aug, 0)
                    kps_aug.append(kp.deepcopy(x=med[0], y=med[1]))

            result[i] = kpsoi.deepcopy(keypoints=kps_aug)

        return result

    def _augment_polygons(self, polygons_on_images, random_state, parents,
                          hooks):

Instances