diff --git a/examples/tracking/visual/PyCV_TrackFeaturesFwdBck.jl b/examples/tracking/visual/PyCV_TrackFeaturesFwdBck.jl
index 3618b53d0..5e28852b3 100644
--- a/examples/tracking/visual/PyCV_TrackFeaturesFwdBck.jl
+++ b/examples/tracking/visual/PyCV_TrackFeaturesFwdBck.jl
@@ -12,6 +12,12 @@ using SHA: sha256
 np = pyimport("numpy")
 cv = pyimport("cv2")
 
+pushfirst!(PyVector(pyimport("sys")."path"), @__DIR__ )
+
+SscPy = pyimport("PySSCFeatures")
+ssc = SscPy."ssc"
+
+
 # # lk_params = ( winSize  = (19, 19), maxLevel = 2, criteria = (cv.TERM_CRITERIA_EPS | cv.TERM_CRITERIA_COUNT, 10, 0.03))
 # lk_params = ( winSize  = (19, 19), maxLevel = 2, criteria = (cv.TERM_CRITERIA_EPS | cv.TERM_CRITERIA_COUNT, 30, 0.01))
 # feature_params = (maxCorners = 1000, qualityLevel = 0.01, minDistance = 8, blockSize = 19 )
@@ -43,16 +49,22 @@ function goodFeaturesToTrack(im1, feature_params; mask=nothing)
   cv.goodFeaturesToTrack(collect(reinterpret(UInt8, im1)), mask=collect(reinterpret(UInt8,mask)), feature_params...)
 end
 
-function goodFeaturesToTrackORB(im1; mask=nothing, orb = cv.ORB_create()) 
+function goodFeaturesToTrackORB(im1; mask=nothing, orb = cv.ORB_create(), downsample::Int=5, tolerance::Real = 0.1) 
   # gray = cv2.cvtColor(im1,cv.COLOR_BGR2GRAY)
   # kypts, decrs = orb.detectAndCompute(gray,None)
   # https://docs.opencv.org/3.4/d1/d89/tutorial_py_orb.html
   # find the keypoints with ORB
   img = collect(reinterpret(UInt8, im1))
   kp = orb.detect(img, collect(reinterpret(UInt8,mask)))
+
+  # downselect a better distribution of features
+  rows, cols = size(img,1), size(img,2)
+  sel_kp = ssc(kp, orb.getMaxFeatures() ÷ downsample, tolerance, cols, rows)
+
   # compute the descriptors with ORB
-  kp, des = orb.compute(img, kp)
-  return kp, des
+  kp_, des = orb.compute(img, sel_kp)
+
+  return kp_, des
 end
 
 function combinePlot(ref_img, overlay_img)
diff --git a/examples/tracking/visual/PySSCFeatures.py b/examples/tracking/visual/PySSCFeatures.py
new file mode 100644
index 000000000..78f57d52a
--- /dev/null
+++ b/examples/tracking/visual/PySSCFeatures.py
@@ -0,0 +1,131 @@
+# Taken from:  https://github.com/BAILOOL/ANMS-Codes/blob/568fa6a1b39aa46fa04ad600cb40a33decf52897/Python/ssc.py#L1
+
+# @article{bailo2018efficient,
+#   title={Efficient adaptive non-maximal suppression algorithms for homogeneous spatial keypoint distribution},
+#   author={Bailo, Oleksandr and Rameau, Francois and Joo, Kyungdon and Park, Jinsun and Bogdan, Oleksandr and Kweon, In So},
+#   journal={Pattern Recognition Letters},
+#   volume={106},
+#   pages={53--60},
+#   year={2018},
+#   publisher={Elsevier}
+# }
+
+# MIT "Expat" License
+
+# Copyright (c) 2018 Oleksandr Bailo
+
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+
+# The above copyright notice and this permission notice shall be included in all
+# copies or substantial portions of the Software.
+
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+# SOFTWARE.
+
+
+import math
+
+def ssc(keypoints, num_ret_points, tolerance, cols, rows):
+    exp1 = rows + cols + 2 * num_ret_points
+    exp2 = (
+        4 * cols
+        + 4 * num_ret_points
+        + 4 * rows * num_ret_points
+        + rows * rows
+        + cols * cols
+        - 2 * rows * cols
+        + 4 * rows * cols * num_ret_points
+    )
+    exp3 = math.sqrt(exp2)
+    exp4 = num_ret_points - 1
+
+    sol1 = -round(float(exp1 + exp3) / exp4)  # first solution
+    sol2 = -round(float(exp1 - exp3) / exp4)  # second solution
+
+    high = (
+        sol1 if (sol1 > sol2) else sol2
+    )  # binary search range initialization with positive solution
+    low = math.floor(math.sqrt(len(keypoints) / num_ret_points))
+
+    prev_width = -1
+    selected_keypoints = []
+    result_list = []
+    result = []
+    complete = False
+    k = num_ret_points
+    k_min = round(k - (k * tolerance))
+    k_max = round(k + (k * tolerance))
+
+    while not complete:
+        width = low + (high - low) / 2
+        if (
+            width == prev_width or low > high
+        ):  # needed to reassure the same radius is not repeated again
+            result_list = result  # return the keypoints from the previous iteration
+            break
+
+        c = width / 2  # initializing Grid
+        num_cell_cols = int(math.floor(cols / c))
+        num_cell_rows = int(math.floor(rows / c))
+        covered_vec = [
+            [False for _ in range(num_cell_cols + 1)] for _ in range(num_cell_rows + 1)
+        ]
+        result = []
+
+        for i in range(len(keypoints)):
+            row = int(
+                math.floor(keypoints[i].pt[1] / c)
+            )  # get position of the cell current point is located at
+            col = int(math.floor(keypoints[i].pt[0] / c))
+            if not covered_vec[row][col]:  # if the cell is not covered
+                result.append(i)
+                # get range which current radius is covering
+                row_min = int(
+                    (row - math.floor(width / c))
+                    if ((row - math.floor(width / c)) >= 0)
+                    else 0
+                )
+                row_max = int(
+                    (row + math.floor(width / c))
+                    if ((row + math.floor(width / c)) <= num_cell_rows)
+                    else num_cell_rows
+                )
+                col_min = int(
+                    (col - math.floor(width / c))
+                    if ((col - math.floor(width / c)) >= 0)
+                    else 0
+                )
+                col_max = int(
+                    (col + math.floor(width / c))
+                    if ((col + math.floor(width / c)) <= num_cell_cols)
+                    else num_cell_cols
+                )
+                for row_to_cover in range(row_min, row_max + 1):
+                    for col_to_cover in range(col_min, col_max + 1):
+                        if not covered_vec[row_to_cover][col_to_cover]:
+                            # cover cells within the square bounding box with width w
+                            covered_vec[row_to_cover][col_to_cover] = True
+
+        if k_min <= len(result) <= k_max:  # solution found
+            result_list = result
+            complete = True
+        elif len(result) < k_min:
+            high = width - 1  # update binary search range
+        else:
+            low = width + 1
+        prev_width = width
+
+    for i in range(len(result_list)):
+        selected_keypoints.append(keypoints[result_list[i]])
+
+    return selected_keypoints