Abstract In this work, we address the problem of segmenting railway tracks from video recordings captured from a train engine, focusing on the rails the train travels on. Our approach is motivated by the physical properties of railway rails, which typically exhibit a distinctive reflective appearance and remain visually and geometrically consistent across consecutive video frames. Moreover, the rails do not undergo significant shape changes or spatial displacement between frames. These properties enable the transfer of contextual information between frames and allow the use of composite models with a reduced number of trainable parameters. We introduce a composite model that incorporates contextual information from previous frames, reflecting the limited temporal variability of rail geometry and appearance, thereby reducing overfitting when training on small annotated datasets. We evaluate three neural network-based segmentation approaches: a convolutional neural network (CNN), a CNN combined with post-processing using morphological transformations and connected component labelling, and the proposed context-aware composite model. Furthermore, we propose an unsupervised evaluation methodology that assesses segmentation quality based on the distinguishability of segment colours from the background, geometric complexity, and colour variance within segments. All models are trained exclusively on synthetic data, with evaluation performed on real-world video sequences. Experimental results demonstrate that incorporating physically interpretable temporal context improves segmentation consistency compared to frame-wise methods. The synthetic dataset created for this work is publicly released.

A new representation for objects with multiple colours-the colour adjacency graph (CAG)-is proposed. Each node of the CAG represents a single chromatic component of the image defined as a set of pixels forming a unimodal cluster in the chromatic scattergram. Edges encode information about adjacency of colour components and their reflectance ratio. The CAG is related to both the histogram and region adjacency graph representations. It is shown to be preserving and combining the best features of these two approaches while avoiding their drawbacks. The proposed approach is tested on a range of difficult object recognition and localisation problems involving complex imagery of non-rigid 3D objects under varied viewing conditions with excellent results.>

The article describes a colour based recognition system with three novel features. Firstly, the proposed system can operate in environments where spectral characteristics of illumination change in both space and time. Secondly, benefits in terms of speed and quality of output are gained by focusing processing to areas of salient colour. Finally, an automatic model acquisition procedure allows rapid creation of the model database. 1 Introduction In the paper we present a colour-based recognition system that aims to demonstrate the advantages of selective processing. We do not attempt to analyse the whole image in the spirit of traditional segmentation methods (eg. [KSK87],[GJT87]); instead we try to find areas where distinctive colour provides least ambiguous information about presence of objects from the model database. Using this approach, standard recognition tasks (eg. What is in the scene?, Where is object X?) can be accomplished without wasting computational resources in parts o...

This paper presents a model-based method for colour-based recognition of objects from a large database. The algorithm is based on the assumption that surface reflectances of objects in the model database follow the extended dichromatic model proposed by Shafer [Sha84]. Adoption of the dichromatic model allows recovery of body colour - the component, of sensor responses (RGB-values) that is independent of scene geometry and illumination intensity. Both theoretical studies [Hea89b] and experiments [LB90][KSK88] confirm that Shafer's model gives a suitable approximation for reflectances of a wide range of materials. Instead of using traditional techniques (eg. clustering, split-and-merge) to obtain regions of 'similarly' coloured pixels followed by classification a novel approach is argued for. First, for each pixel a list of models with nonzero aposteriori probabilities P(modeU\body colotir) is computed using Bayes formula. Next, regions are formed by grouping pixels with identical most probable hypothesis. Probabilities P(modeli\region) are obtained trough a standard group decision rule [FT80]. We show that the proposed scheme can be used for a number of visual tasks - localization of objects, generation and verification of object hypotheses. Experiments on images of complex indoor scenes confirm that the proposed method can provide reliable information about the surrounding environment.