Fast range image segmentation using high-level segmentation primitives

We present fast and accurate segmentation algorithms of range images of urban scenes. The utilization of these algorithms is essential as a pre-processing step for a variety of tasks, that include 3D modeling, registration, or object recognition. The accuracy of the segmentation module is critical for the performance of these higher-level tasks. In this paper, we present a novel algorithm for extracting planar, smooth non-planar, and non-smooth connected segments. In addition to segmenting each individual range image, our methods also merge registered segmented images. That results in coherent segments that correspond to urban objects (such as facades, windows, ceilings, etc.) of a complete large scale urban scene. We present results from experiments of one exterior scene (Cooper Union building, NYC) and one interior scene (Grand Central Station, NYC).

In this paper, a fast watershed-based image segmentation algorithm is proposed. Firstly, a homogeneity gradient image named after H-image is produced, and then the traditional watershed segmentation is used to initially segment the H-image to produce over-segmented image regions (objects), and at last local optimum merging strategy is utilized to merge neighbor regions to construct a hierarchical network of image objects according to different scales. Textural, color and shape information of segments is utilized in the merging process. Experiments demonstrate the validity of the algorithm proposed in this paper.

This paper considers the segmentation of range image measurements into surface patches which are either plane or curved and which are described formally by a function. After a formal description of the segmentation, we present and compare three methods suited for plane and curved patch segmentation and show the results of experiments conducted for testing their practical behaviour. The two first methods use the classical approach of region growing whereas the third method is based on a relaxation process. This original and last method exhibits simplicity and low computational complexity. Thanks to its parallel nature, it can be considered as a good candidate for range image segmentation in real-time applications.

Distributed and parallel computing techniques allow fast image processing, namely when these techniques are applied at the low and the medium level of a vision system. In this paper, a collective and distributed method for image segmentation is introduced and evaluated. The method is modeled as a multi-agent system, where the agents aim to collectively produce a region-based segmentation. Each agent starts searching for an acceptable region seed by randomly jumping within the image. Next, it performs a region growing around its position. Thus, several agents find themselves within the same homogeneous region and are organized in a graph where two agents are connected if they are within the same region. So, a unifying of the labels in a same region is collaboratively performed by the agents themselves. The proposed method was experimented on real range images from the ABW dataset and the Object Segmentation Database (OSD) one, and the obtained results were compared to those of some well-referenced methods from the literature. The evaluation results show that the proposed method provides fast and accurate image segmentation, allowing it to be deployed for real-time vision systems.

A robust active contour model driven by pre-fitting bias correction and optimized fuzzy c-means algorithm for fast image segmentation

A technique to generate a photorealistic three-dimensional (3D) image and color-textured model using a dedicated optical digitizer is presented. The proposed technique is started with the range and texture image acquisition from different viewpoints, followed by the registration and integration of multiple range images to get a complete and nonredundant point cloud that represents a real-life object. The accuracy of the range image and the precision of correspondence between the range image and texture image are guaranteed by sensor system calibration. Based on the point cloud, a geometric model is established by considering the connectivity of adjacent range image points. In order to enhance the photorealistic effect, we suggest a texture blending technique that utilizes a composite-weight strategy to blend the texture images within the overlapped region. This technique allows more efficient removal of the artifacts existing in the registered texture image, leading to a 3D image with photorealistic quality and color-texture modeling. Experimental results are also presented to testify to the validity of the proposed method.

Multi-view range image registration is a significant and challenging problem for 3D modeling. This paper presents a reference shape based multi-view range image registration algorithm. First, a set of Rotational Projection Statistics (RoPS) features are extracted from the input range images. Next, the reference shape is initialized by selecting a range image from the input. The reference shape is then iteratively updated by registering itself with the remaining range images. The registration between the reference shape and any range image is completed by RoPS feature matching. Finally, all input range images are registered according to their corresponding reference shapes. A number of experiments were performed to test the performance of our algorithm. The experimental results show that the reference shape based algorithm can perform multi-view registration on a mixed set of unordered range images corresponding to several different objects. It is also very accurate and efficient. It outperformed the state-of-the-arts including the spanning tree based and connected graph based algorithms.

High-level vision systems that use range images require them to be segmented into clusters of pixels with a common geometric representation. This paper presents a new range image segmentation technique based on successive differentiation. Each time the image is differentiated, an edge detection process follows to locate the surface jump and fold discontinuities. Then, the closed edge contours and the different image regions are identified. The segmentation process is relatively fast and requires minimal user-interference. Experimental results using real range images are presented.

An adaptive range image coding algorithm for the geometry compression of large-scale 3D point clouds (LS3DPCs) is proposed in this work. A terrestrial laser scanner generates an LS3DPC by measuring the radial distances of objects in a real world scene, which can be mapped into a range image. In general, the range image exhibits different characteristics from an ordinary luminance or color image, and thus the conventional image coding techniques are not suitable for the range image coding. We propose a hybrid range image coding algorithm, which predicts the radial distance of each pixel using previously encoded neighbors adaptively in one of three coordinate domains: range image domain, height image domain, and 3D domain. We first partition an input range image into blocks of various sizes. For each block, we apply multiple prediction modes in the three domains and compute their rate-distortion costs. Then, we perform the prediction of all pixels using the optimal mode and encode the resulting prediction residuals. Experimental results show that the proposed algorithm provides significantly better compression performance on various range images than the conventional image or video coding techniques

Based on the spherical harmonics representation of image formation, we derive a new photometric metric for evaluating the correctness of a given rigid transformation aligning two overlapping range images captured under unknown, distant, and general illumination. We estimate the surrounding illumination and albedo values of points of the two range images from the point correspondences induced by the input transformation. We then synthesize the color of both range images using albedo values transferred using the point correspondences to compute the photometric reprojection error. This way allows us to accurately register two range images by finding the transformation that minimizes the photometric reprojection error. We also propose a practical method using the proposed photometric metric to register pairs of range images devoid of salient geometric features, captured under unknown lighting. Our method uses a hypothesize-and-test strategy to search for the transformation that minimizes our photometric metric. Transformation candidates are efficiently generated by employing the spherical representation of each range image. Experimental results using both synthetic and real data demonstrate the usefulness of the proposed metric.

Time-of-flight range imaging cameras measure distance to objects in their field of view, but are prone to error when objects move. At least three raw frames are required to obtain one range image, and the standard method is to read out raw frames into separate sets and process to find one range image per set. Motion during the acquisition of a set causes error in the corresponding range image. In this paper, the problem of motion is addressed by regarding the raw data from each pixel as a noisy time series, and using the Kalman filter to efficiently perform time-series analysis. The proposed method adapts to the effects of transverse motion, measuring a sharp range image at each raw frame. The error in the proposed method is less than the traditional approach in 80% of tests, with no detected increase in the STD due to noise. In the qualitative experimental results, the visible blur is reduced.

Edge detection in range images provides rich information about the scene being observed. Many researchers have been taking advantage of this information to improve image segmentation by integrating edge detection with other segmentation techniques. This paper presents a methodology to perform edge detection in range images in order to provide a reliable and meaningful edge map, which helps to guide and improve range image segmentation by clustering techniques. The obtained edge map leads to three important improvements: (1) the definition of the ideal number of clusters to initialise the clustering algorithm; (2) the selection of suitable initial cluster centres; and (3) the successful identification of distinct regions with similar features. Experimental results that substantiate the effectiveness of this work are presented.

A versatile full-field range imaging system has previously been constructed. This system is configurable in software to produce either high precision or fast acquisition range images. Indicatively a 10 second exposure has been shown to produce a range image of sub-millimeter precision, whilst video frame rate (30 fps) acquisition provides for centimetre precision. Currently the acquisition time of the system is to a large degree constrained by the off-line processing of the frames by an external computer. This paper presents an alternative to the off-line PC image processing utilising an Altera Stratix II FPGA. Processing rates up to 30 frames per second have been achieved with the added advantage that many of the previous systempsilas existing digital electronics can also be accommodated, providing for an even more compact and flexible system.

A new algorithm for planar face segmentation in range images is introduced. The segmentation is based on a combination of recursive adaptive regression model prediction for detecting range image step discontinuities which are present at object face borders and of a region growing on surface lines. Border pixels are detected in two perpendicular directions and detection results are combined together. Two predictors in each direction use identical contextual information from the pixel neighbourhood and they mutually compete for the most optimal discontinuity defection.

Image segmentation is a crucial step in any image analysis process. It consists in preparing the image to make it more usable by an automatic process, such interpreting and understanding its content. In this paper, we introduce a novel method for range image segmentation. The proposed method proceeds by adapting the principle of the Canny filter, commonly utilized for greyscale and colour images, to be applied for range images, where depths in these images should be differently handled. Instead of using raw image data, a new image of angles between normal vectors according some given directions is computed and then used to compute a new gradient image. Canny steps are then applied on the latter image, producing precise and well located edges. Experimentation on real images from the ABW database shows that edges in range images are correctly detected.