SYSTEMS, CIRCUITS, AND METHODS FOR EFFICIENT HIERARCHICAL OBJECT RECOGNITION BASED ON CLUSTERED INVARIANT FEATURES

Abstract

One embodiment is a method for selecting and grouping key points extracted by applying a feature detector on a scene being analyzed. The method includes grouping the extracted key points into clusters that enforce a geometric relation between members of a cluster, scoring and sorting the clusters, identifying and discarding clusters that are comprised of points which represent the background noise of the image, and sub-sampling the remaining clusters to provide a smaller number of key points for the scene.

Description

TECHNICAL FIELD

Embodiments of the present invention relate generally to visual search systems and, more specifically to systems, circuits, and methods that group feature descriptors of a captured scene into clusters to improve matching with reference images and the efficiency of transmission of such feature descriptors.

BACKGROUND

Visual search systems are known and operate to use captured images as “queries” for a database in order to retrieve information related to the content of the captured image. For example, after taking a photo of the facade of a museum, a user's smartphone that was used to capture the image processes the image to generate feature descriptors that effectively describe the image for purposes of the query. In such a situation, which is a rapidly growing area of visual search, the smartphone thereafter communicates the generated feature descriptors to a remote system containing the database, and searches or queries the database using the feature descriptors to identify the captured image in the database. The remote system may thereafter communicate to the smartphone the results of this query for presentation to the user, such as the location, opening times, and ticket costs where the captured image is a museum that the user is interested in seeing.

A typical visual search pipeline includes an interest points detector, a features descriptor generator, a matching stage, and a geometry consistency checker. The most successful visual search techniques make use of invariant features, where the term invariant refers to the ability of the detection algorithm to detect image points and describe its surrounding region in order to be tolerant to affine transformation, like rotation, translation and scaling. In the literature there exist many invariant features extraction and description algorithms. The Scale Invariant Feature Transform (SIFT) algorithm is often used as a reference algorithm because it is typically the algorithm that provides the best recognition rate (i.e., matching of a descriptors associated with a captured image with the proper corresponding image in an image database). The computational costs of the SIFT algorithm, however, are quite high (e.g., less than 2 frames per second (FPS) at VGA resolution on a modern desktop computer) so there are other algorithms, like the Speeded Up Robust Features (SURF) algorithm, that sacrifice precision for speed.

In the typical visual search pipeline, the matching stage simply consists of a many to many comparison between the descriptors in the scene and the ones stored in the database using a predefined metric function which is dependent to the description algorithm (e.g. ratio of L2 norms for SIFT). Finally, the geometry consistency checker, such as a Random Sample Consensus (RANSAC) checker, processes the entire set of matched features to retrieve a valid transformation model for the query and reference images with the goal of removing false matches (outliers) and increasing the recognition quality. Typical visual search applications are pair-wise matching, which simply compares two images, and a more complex analysis named retrieval, in which a query image is looked up inside a reference image data set, potentially very huge like Google Images or Flickr.

Each of the phases described above requires high computational costs due to the amount of data involved or the complexity of the calculations. Also the amount of byte used to the descriptor is an important factor for a potential transmission overhead in a client-server environment, such where a mobile device like a smartphone is the image capture device, as well as for amount of storage space required to store all the desired the content of the database. Improvements in the matching stage have been proposed, such as by the creator of the SIFT algorithm who proposed the use of KD-Trees to approximate the lookup of the nearest feature vector inside a database. Other improvements have been applied to the geometry consistency checker using a algorithms that are less computationally intensive than RANSAC, such as the DISTRAT algorithm. Finally the compression and transmission of the image's features in the form of the feature descriptors is the core topic of an MPEG standardization group named Compact Descriptors for Visual Search (CDVS). For example, the Compress Histogram of Gradients (CHOG) is a feature descriptor algorithm or approach designed to produce a compact representation by applying sampling and discretization to SIFT-like feature descriptors.

SUMMARY

An embodiment is a method for selecting and grouping key points extracted by applying a feature detector on a scene being analyzed. The method includes grouping the extracted key points into clusters that enforce a geometric relation between members of a cluster, scoring and sorting the clusters, identifying and discarding clusters that are comprised of points which represent the background noise of the image, and sub-sampling the remaining clusters to provide a smaller number of key points for the scene. The method of can include iteratively applying the operations of grouping through sub-sampling to a set of reference images including sets comprised of a single reference image. The method can further include using clusters information to match clusters in the two images by using a metric which is defined by selected invariant features descriptors and by exploiting the sub-sampled cluster data to reduce the complexity of the matching phase including an early discard of clusters that cannot be probably found between extracted clusters. The method also can include tracking multiple independently moving objects in the scene being analyzed or detecting multiple objects found in the scene being analyzed.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is an image illustrating clustering of descriptors in an image according to one embodiment described in the present disclosure.

FIG. 2 is an image illustrating matching between images having different scales.

FIG. 3 is a functional block diagram illustrating a sequential visual search pipeline according to one embodiment described in the present disclosure.

FIG. 4 is a functional block diagram illustrating a parallel visual search pipeline according to another embodiment described in the present disclosure.

FIG. 5 is a functional block diagram of the clustering module of FIGS. 3 and 4 according to one embodiment described in the present disclosure.

FIG. 6 is a functional block diagram of visual search system including the pipelines of FIG. 3 or 4 according to one embodiment of the present disclosure.

DETAILED DESCRIPTION

Visual search for mobile devices relies on transmitting wirelessly a compact representation of the query image, generally in the form of feature descriptors, to a remote server. Descriptors are therefore compressed, so as to reduce bandwidth occupancy and network latency. Given the impressive pace of growth of 3D video technology, 3D visual search applications for the mobile and the robotic market will become a reality. Accordingly, embodiments described herein are directed to improving detection and reducing the bandwidth required for such prospective applications through a hierarchical method using clustering of invariant features of a scene being analyzed.

A representative visual search system 100 is illustrated in FIG. 6 and includes local image capture device 102, such as a mobile device like a smart phone, and automobile including two-dimensional or three-dimensional sensors for navigation, or an image capture system positioned at a certain location in a city such as a camera mounted on top of a lamppost at a particular intersection. The local image capture device 102 generates two- or three-dimensional uncompressed feature descriptors for the scene being imaged. These descriptors must then be communicated over a communications network 104 to a remote server 106 containing a visual database that will be queried to identify an image corresponding to the local image captured by the device 102 and represented by the descriptors. The remote server 106 then returns the visual search results to the device 102 for use by the device or a user of the device. Embodiments disclosed herein are directed to methods of compressing these uncompressed feature descriptors so that the bandwidth of the communications network 104 is sufficient to provide the desired operation of the system 100. As illustrated in FIG. 6, the device 102 provides compressed feature descriptors over the communication network 104 to the remote server 106.

In the following description, certain details are set forth to provide a sufficient understanding of the present invention, but one skilled in the art will appreciate that the invention may be practiced without these particular details. Furthermore, one skilled in the art will appreciate that the example embodiments described below do not limit the scope of the present invention, and will also understand various modifications, equivalents, and combinations of the disclosed example embodiments and components of such embodiments are within the scope of the present invention. Illustrations of the various embodiments, when presented by way of illustrative examples, are intended only to further illustrate certain details of the various embodiments, and should not be interpreted as limiting the scope of the present invention. Finally, in other instances below, the operation of well known components, processes, algorithms and protocols have not been shown or described in detail to avoid unnecessarily obscuring the present invention.

State of art invariant features extractors and descriptors initially produce a large set of features and feature descriptors. After the matching and geometry checking phases, the number of features that could be correctly paired are an order of magnitude less. Because the matching stage compares each feature in query image A with all the features in a reference image B, it means that the matching stage wastes a lot of time computing useless correspondences. In the literature some improvement are presented to improve comparison time performance by applying, for example, approximated research strategy based on trees.

Moreover a query image doesn't usually contain only the subject of the research but also other objects and a background environment that introduce noise in the query set. Embodiments described herein group the identified interest points by evaluating a geometric constraint between them in order to later apply a full search comparison to smaller features data set which are geometrically similar. Embodiments also automatically identify and remove noisy features in the query image and send the smallest number possible of features to the matching phase (i.e., from the device 102 to remoter server 106 in FIG. 6). Embodiments also detect multiple objects in a scene being analyzed.

The first step in embodiments described herein, after key point detection or feature extraction to generate corresponding feature descriptors, is a clustering stage that groups the points or feature descriptors using a desired metric. In one embodiment, the metric exploits the spatial coherence of the points or feature descriptors.

In one embodiment, the metric computes the distance of each point or feature descriptor from its neighbors in applying an 8-distance methodology in a prefixed circular region. If memory is available and a fixed image size is being used, this step can be replaced with a look-up table to save computational cost. The metric then keeps creating and/or merging clusters following a proximity relation. The metric or algorithm keeps running until the size and density of the clusters reach a variable threshold. This process builds a variable number of clusters with different densities that we can associate to the idea of the quantity of information of the area. In accordance with this interpretation it is then possible to remove the low density sets or the smallest ones in order to focus the computation only on the portion of the image which is rich in information (i.e., contains a lot of feature descriptors).

This approach removes background and low density information regions that hardly produce a correct association at the end of the entire process. This also allows ranking between several features clusters that can be exploited by the matching strategy, and potentially allows the algorithm to work only on the most significant image region. An example output of this section of the system is shown in FIG. 1. The top image in FIG. 1 displays the interest points detected by the SIFT algorithm, while the bottom image represents the clusters identified by this embodiment. The left side of both images (highlighted in blue, portion to the left of the vertical line containing the hand) shows a clear example of the automated noise-removal. Both the background and the hand are detected as “foreign” to the scene and most of their interest points or feature descriptors are correctly discarded.

FIG. 5 is a block diagram of the visual search pipeline up to this point. Normally the reference image portrays the object without ambiguity, so the top left edge of the object lie exactly in the same position in the image. This assumption is not true for the query image, because the object can appear in any position or rotation, preventing the trivial match of the top left region of the two images. The proposed solution consists of randomly selecting, in each features cluster of both images, a number of evenly spread candidates and performing a full matching on this data. This procedure allows the removal of reference system incoherence and quickly discriminating different objects using only a restricted number of features. In tests objects were successfully identified using only 20% of the initial data.

Another issue is introduced by possible difference of scale of an object between the reference and the query image. The increment of information in the magnified image, in fact, can generate multiple clusters that have to be potentially associated with a single cluster in the other image. To avoid this incoherence the proposed pipeline merges clusters that are part of a many-to-one relation. FIG. 2 shows a sample result of this phase.

Finally, to refine this quick discrimination and obtain an unique and robust result, the matching stage requests the transmission of the features belonging to the survived sets and after a new matching process, using all available data, removes most of the outliers using a RANSAC strategy to each cluster association.

The process explained in the embodiments above allows multiple configurations of the pipeline that have to be chosen with respect to the target application. A first possible configuration, namely a sequential pipeline, is shown in FIG. 3 and waits for the quick matching response and then computes the descriptors only in the survived sets potentially saving a lot of computation, in particular in the case of a non-matching query image. This obviously leads to a better power consumption since the processor performing these operations can idle during the matching stage if that is done on another system.

A second configuration shown in FIG. 4 is a parallel pipeline configuration and starts the computation of the feature descriptors of the remaining sets while the matching stage is still processing the quick matches. Despite the increment in power consumption due to the possible wasted computation, this approach is useful if we are looking for a fast when the search phase is slow (e.g. slow network transfer speed or a big data base)

Since the algorithm applies the geometry consistency check on the clusters pairs instead of the whole matching descriptors of the scene, embodiments according to the disclosed approach allow the recognition of objects belonging to two different scenes when no real affine transformation exists between the scene themselves (e.g. the system can track two objects that are moving independently one from the other).

One skilled in the art will understand that even though various embodiments and advantages of the present invention have been set forth in the foregoing description, the above disclosure is illustrative only, and changes may be made in detail, and yet remain within the broad principles of the invention. For example, many of the components described above may be implemented using either digital or analog circuitry, or a combination of both, and also, where appropriate, may be realized through software executing on suitable processing circuitry. It should also be noted that the functions performed can be combined to be performed by fewer elements or process steps depending upon the actual embodiment being used in the system 100 of FIG. 1. Therefore, the present invention is to be limited only by the appended claims.

Claims

1. A method for selecting and grouping key points extracted by applying a feature detector on a scene being analysed, the method comprising: grouping the extracted key points into clusters that enforce a geometric relation between members of a cluster;scoring and sorting the clusters;identifying and discarding clusters that are comprised of points which represent the background noise of the image; andsub-sampling the remaining clusters to provide a smaller number of key points for the scene.

2. The method of claim 1 comprising iteratively applying the operations of grouping through sub-sampling to a set of reference images including sets comprised of a single reference image.

3. The method of claim 2 further comprising using clusters information to match clusters in the two images by using a metric which is defined by selected invariant features descriptors and by exploiting the sub-sampled cluster data to reduce the complexity of the matching phase including an early discard of clusters that cannot be probably found between extracted clusters.

4. The method of claim 3 further comprising tracking multiple independently moving objects in the scene being analyzed.

5. The method of claim 3 further comprising detecting multiple objects found in the scene being analyzed.