DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
2. This is in response to the applicant response filed on 09/03/2026. In the applicant’s response, claims 1, 7, and 12-13 were amended; claims 5-6, 9, 11, and 17 were cancelled. Accordingly, claims 1-4, 7-8, 10, 12-16, and 18-20 are pending and being examined. Claims 1 and 7 are independent form.
Claim Rejections - 35 USC § 101
3. The claims rejected under 35 USC § 101 made in the previous office action have been withdrawn in view of applicant’s amendment.
Claim Rejections - 35 USC § 103
4. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
5. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
6. Claims 1-4, 7-8, 10, 12-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Dersch et al (“Combining graph-cut clustering with object-based stem detection for tree segmentation in highly dense airborne lidar point clouds”, 2021, hereinafter “Dersch”) in view of Li et al (“A New Method for Segmenting Individual Trees from the Lidar Point Cloud”, 2012, hereinafter “Li”).
Regarding claim 1, Dersch discloses a forestry management system for identifying one or more characteristics of a tree from point cloud data obtained from a laser scan or photographic image of one or more trees, the forestry management system comprising: a processor configured with hardware and/or software to execute steps to (the tree detection system; see fig.1, the title and abstract);
input the point cloud data (“highly dense 3D point cloud”; see fig.1; see fig.7 (a)),
segment the tree from the point cloud data using unsupervised, graph-based clustering (“single tree segmentation” by the “graph-cut based clustering”; see fig.1 and Sec. 3.3. It should be noticed that the graph-cut based clustering is an unsupervised, i.e., based on k-means clustering),
identify a metric of the tree using an algorithm, determine a trunk location of the tree (see sec. 6.3, paragraph 2, lines 7-11; “our procedure provides not only the segmented point cloud per tree and important tree parameter (e.g. position and height of the tree), but also the tree trunk as 3D information, which in turn can be used to model the tree trunk as an object.”)
As explained and interpreted above, the mere difference is, Dersch does not explicitly disclose wherein the tree detection system is based “in a canopy-to-root routing direction that identifies at least-cost route directed from canopy points toward a ground point at a base of the tree and simultaneously segments the point cloud data into individual trees and discovers stem locations of the trees in a single operation” as recited by claim 1. However, in the same field of endeavor, Li teaches a tree segmentation which, first normalizes the point cloud values by subtracting the ground points from the lidar point cloud to determine the height of the tree (see fig.2 and “Preprocessing” and “The Algorithm”, on the right col., on page 76), and then “isolates trees individually and sequentially from the point cloud, from the tallest tree to the shortest” by adopting “a top-to-bottom approach to classify the points, i.e., classifying the points in Ui one by one, starting from the highest point to the lowest one. See “Implementation”, on the right col., on page 78; see fig.2—fig.4. It would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention was made to incorporate the teachings of Li into the teachings of Dersch and adopt the top-to-bottom approach of Li to segment trees from point cloud. Suggestion or motivation for doing so would have been to segment individual trees from the Lidar point cloud as taught by Li, see, the title. Therefore, claim 1 is unpatentable over Dersch in view of Li.
Regarding claim 2, the combination of Dersch and Li discloses the forestry management system of Claim 1, wherein the metrics include at least one of a height of the tree, a biomass of the tree, a health status of the tree, and a species of the tree (Dersch, see sec. 1, para.1).
Regarding claim 3, the combination of Dersch and Li discloses the forestry management system of Claim 1, wherein the metric is a stem location and a position of the tree (Dersch, e.g., see the tree stem’s locations shown in fig.7(d)).
Regarding claim 4, the combination of Dersch and Li discloses the forestry management system of Claim 3, wherein the position of the tree is the angle of a trunk of the tree in relation to a ground surface (Dersch, see “the fitted magenta lines” in fig.7(d)).
Regarding claim 7, Dersch method of identifying one or more characteristics of a tree from point cloud data obtained from a laser scan or photographic image of one or more trees (the tree detection system from lidar point cloud; see fig.1, the title and abstract), the method comprising the steps of:
preprocessing the point cloud data by identifying a plurality of voxel cells and aggregating points within a corresponding voxel cell, thus forming a superpoint from each aggregation (see the paragraph right below the Equation (11): from the lidar point cloud, “[w]e subdivided the area of interest into a voxel space, which resulted in equal-sized primitives {Oij} in the form of voxels with a side length of 0.5 m.” It should be noticed: wherein each of N primitives {Oij}, i=1, 2, ...N, is a super-voxel or super-point. See the top paragraph in the left col. on page. 212.);
building a graph model from the superpoints (see the paragraph right below the 3.3—“Graph-cut cluster”: creating a weighted graph G= (E, V) with V as the node set, E as the edge set, wij(oi, oj) as W ={wij(oi, oj)}i=1…N, j=1…N as the similarity matrix representing the pairwise weighted interrelationship between N primitives O = {oi}, i=1…N of a set of cubic voxels or super-voxels); identifying ground superpoints and canopy superpoints from heights of the superpoints (see fig.7(a), wherein colors are by height from the ground);
see sec. 6.3, paragraph 2, lines 7-11; “our procedure provides not only the segmented point cloud per tree and important tree parameter (e.g. position and height of the tree), but also the tree trunk as 3D information, which in turn can be used to model the tree trunk as an object.”).
As explained and interpreted above, the mere difference is, Dersch does not explicitly disclose wherein the tree detection system is based in “identifying the canopy to root path of the tree using a least-cost routing routine that identifies the least-cost route from canopy superpoints down to a ground superpoint at a base of the tree” as recited by claim 7. However, in the same field of endeavor, Li teaches a tree segmentation which, first normalizes the point cloud values by subtracting the ground points from the lidar point cloud to determine the height of the tree (see fig.2 and “Preprocessing” and “The Algorithm”, on the right col., on page 76), and then “isolates trees individually and sequentially from the point cloud, from the tallest tree to the shortest” by adopting “a top-to-bottom approach to classify the points, i.e., classifying the points in Ui one by one, starting from the highest point to the lowest one. See “Implementation”, on the right col., on page 78; see fig.2—fig.4. It would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention was made to incorporate the teachings of Li into the teachings of Dersch and adopt the top-to-bottom approach of Li to segment trees from point cloud. Suggestion or motivation for doing so would have been to segment individual trees from the Lidar point cloud as taught by Li, see, the title. Therefore, claim 7 is unpatentable over Dersch in view of Li.
Regarding claim 8, the combination of Dersch and Li discloses the method of Claim 7, wherein the step of preprocessing the point cloud data includes normalizing the point cloud data by subtracting a terrain elevation from each point in the point cloud data (Dersch, see the paragraph right above Sec. 3.4: “the implemented solution of the graph-cut clustering method subtracts ground points from the lidar point cloud using an appropriate filtering procedure for a DTM to avoid false ground clusters.”).
Regarding claim 10, the combination of Dersch and Li discloses the method of Claim 9, wherein the step of preprocessing the point cloud data includes applying a point count threshold by ignoring voxels containing fewer than a predetermined number of points (Dersch, see Sec. 3.1, par.1, lines 5-7: “Based on our experiences, the point density must be such that at least five points/m represent the tree trunk.”).
Regarding claim 12, the combination of Dersch and Li discloses the claimed invention. See Dersch, Sec. 6.1, para.2: “Second, the stem detection does not need a DTM to cancel ground points. Fig. 7b shows that most of the ground points are classified in a green and blueish color corresponding to a low point probability ppt. Because we optimize the control parameter pptthres in a sensitivity analysis (see Section 5.3), we discard all points whose point probabilities are below the optimized values pptthres = 0.44. This procedure eliminates most of the irrelevant non-stem points. If we take a look at Fig. 7c, we can notice that several segments (in red) have been filtered out. These segments result from ground points that still remain in the point cloud after elimination using pptthres.” It should be noticed that the pptthres here is a threshold for classifying point cloud to the non-ground points including leaves (i.e., the canopy/crown points showm by fig.7(b)) or the ground points showm by figs.(b)-(d). In other words, if a point belongs to the stem/canopy point, its point probability must be upper than or equal to the pptthres. Likewise, If a point belongs to the groung point, its point probability must be below the pptthres.)
Regarding claim 13, the combination of Dersch and Li discloses the method of Claim 11, wherein the graph model is built by defining an edge between at least two superpoints (Dersch, see Sec. 3.3—Graph-cut clustering: “a weighted graph G = G(E,V) is created with V as the node set, E as the edge set, wij(oi, oj) as the symmetric, non-negative pairwise object similarity function and W ={wij(oi, oj)}, i=1…N,j=1…N as the similarity matrix representing the pairwise weighted interrelationship between N primitives O = {oi}i=1…N of a set of cubic voxels or super-voxels,...”).
Regarding claim 14, the combination of Dersch and Li discloses the method of Claim 13, wherein the graph model is calculated from the algorithm as: N= (P, E, W) (ibid.)
Regarding claim 15, the combination of Dersch and Li discloses the method of Claim 13, wherein the step of identifying the canopy to root path of the tree further includes identifying a cost value of the edge to the ground point (Dersch, see Sec. 3.3, wherein the normalized graph-cut clustering algorithm includes minimizing the objective/cost function defined by Eq(5), which includes calculating the volume of subgraph A that is equivalent to the sum of weights of all edges ending up in cluster A, maximizes the similarity within clusters (vol(A) and vol(B)), and minimizes the similarity between the disjoint clusters A and B (cut(A, B)). It should be noticed that clusters A and cluster B are interpreted as the non-ground points including stem point and leaves (i.e., the canopy) and ground points as shown by fig.7(b).).
Regarding claim 16, the combination of Dersch and Li discloses or suggests the claimed invention. Dersch discloses the similarity/distance function defined by Eq.(8), which includes calculating the Euclidian distance dk(Oi, Oj). It would have been obvious to one of ordinary skill in the art before the effective filling date of the claimed invention was made to appreciate that claim 16 is an obvious variation of Eq(8) and have been equally interchangeable between the similarity function defined by Eq.(8) in Dersch and the similarity function defined by claim 16. Suggestion or motivation for doing so would have been to combine the graph-cut clustering with object-based stem detection for tree segmentation as taught by Dersch, see, Abstract. Therefore, claim 16 is unpatentable over Dersch in view of Li.
Regarding claim 18, the combination of Dersch and Li discloses the method of Claim 7, wherein the step of identifying the canopy to root path of the tree and the step of segmenting the tree from remaining point cloud data occur simultaneously (ibid.).
Regarding claim 19, the combination of Dersch and Li discloses the method of Claim 7, wherein the step of segmenting the tree from remaining point cloud data precedes the step of determining the trunk location of the tree (Dersch, see Abstract, lines 3-8: “This paper describes a novel integrated single tree segmentation using a graph-cut clustering method that is supported by automatic stem detection. The key idea is to replace the static stopping criterion, which is usually defined by trial and error or by a sensitivity analysis, here with a query for whether a stem position has been provided by the stem detection in the remaining cluster to be partitioned. The stem detection automatically detects tree stems by identifying vertical lines based on a hierarchical classification procedure.” In other words, the single tree segmentation result in the method in Dersch is based on the stem position which has been detected by the method).
Regarding claim 20, the combination of Dersch and Li discloses the method of Claim 7, wherein the step of segmenting the tree from remaining point cloud data and the step of determining the trunk location of the tree occur simultaneously (Dersch, see Sec. 3.3, wherein the normalized graph-cut clustering algorithm includes minimizing the objective/cost function defined by Eq(5), which includes calculating the volume of subgraph A that is equivalent to the sum of weights of all edges ending up in cluster A, maximizes the similarity within clusters (vol(A) and vol(B)), and simultaneously minimizes the similarity between the disjoint clusters A (i.e., the canopy) and B (i.e., the ground) (cut(A, B)). It should be noticed that clusters A and cluster B are interpreted as the non-ground points including stem point and leaves (i.e., the canopy) and ground points as shown by fig.7(b).).
Response to Arguments
7. Applicant's arguments with respect to claims 1 and 7 have been considered but are moot in view of the new ground(s) of rejection. As explained in the rejections of the claims, Li discloses or suggests determining a trunk location of a tree in a canopy-to-root routing direction that identifies a least-cost route directed from canopy
points toward a ground point at a base of the tree and simultaneously segments the
point cloud data into individual trees and discovers stem locations of the trees in a
single operation, as recited in claim 1, and identifying the canopy to root path of the tree
using a least-cost routing routine that identifies the least-cost route from canopy
superpoints down to a ground superpoint at a base of the tree, as recited in claim 9.
Conclusion
8. Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
9. Any inquiry concerning this communication or earlier communications from the examiner should be directed to RUIPING LI whose telephone number is (571)270-3376. The examiner can normally be reached 8:30am--5:30pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, HENOK SHIFERAW can be reached on (571)272-4637. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit https://patentcenter.uspto.gov; https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center, and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/RUIPING LI/Primary Examiner, Ph.D., Art Unit 2676
Prosecution Insights