Low-level Texture Features Research Articles

A Hybrid Scene Text Script Identification Network for Regional Indian Languages

In this work, we introduce WAFFNet, an attention-centric feature fusion architecture tailored for word-level multi-lingual scene text script identification. Motivated by the limitations of traditional approaches that rely exclusively on feature-based methods or deep learning strategies, our approach amalgamates statistical and deep features to bridge the gap. At the core of WAFFNet, we utilized the merits of Local Binary Pattern—a prominent descriptor capturing low-level texture features with high-dimensional, semantically-rich convolutional features. This fusion is judiciously augmented by a spatial attention mechanism, ensuring targeted emphasis on semantically critical regions of the input image. To address the class imbalance problem in multi-class classification scenarios, we employed a weighted objective function. This not only regularizes the learning process but also addresses the class imbalance problem. The architectural integrity of WAFFNet is preserved through an end-to-end training paradigm, leveraging transfer learning to expedite convergence and optimize performance metrics. Considering the under-representation of regional Indian languages in current datasets, we meticulously curated IIITG-STLI2023, a comprehensive dataset encapsulating English alongside six under-represented Indian languages: Hindi, Kannada, Malayalam, Telugu, Bengali, and Manipuri. Rigorous evaluation of the IIITG-STLI2023, as well as the established MLe2e and SIW-13 datasets, underscores WAFFNet’s supremacy over both traditional feature-engineering approaches as well as state-of-the-art deep learning frameworks. Thus, the proposed WAFFNet framework offers a robust and effective solution for language identification in scene text images.

S[formula omitted]: Self-supervised learning with sparse-dense sampling

Self-supervised visual representation learning (SSL) attempts to extract significant features from unlabeled datasets, alleviating the necessity for labor-intensive and time-consuming manual labeling processes. However, existing contrastive learning-based methods typically suffer from the underutilization of datasets, consume significant computational resources, and employ longer training epochs or large batch sizes. In this study, we propose a novel method aimed at optimizing self-supervised learning that integrates the advantages of sparse-dense sampling and collaborative optimization, thereby significantly improving the performance of downstream tasks. Specifically, sparse-dense sampling primarily focuses on high-level semantic features, while leveraging the spatial structure relationship provided by the unlabeled dataset to ensure the incorporation of low-level texture features to improve data utilization. Besides, collaborative optimization, including contrastive and location tasks, further enhances the model’s ability to perceive features of different dimensions, thereby improving its utilization of features in the embedding space. Furthermore, the combination of sparse-dense sampling and collaborative optimization strategies can reduce computational consumption while improving performance. Extensive experiments demonstrate that the proposed method effectively reduces the computational requirements while delivering favorable results. The codes and model weights will be available at https://github.com/AI-TYQ/S4.

A Fabric Defect Segmentation Model Based on Improved Swin-Unet with Gabor Filter

Fabric inspection is critical in fabric manufacturing. Automatic detection of fabric defects in the textile industry has always been an important research field. Previously, manual visual inspection was commonly used; however, there were drawbacks such as high labor costs, slow detection speed, and high error rates. Recently, many defect detection methods based on deep learning have been proposed. However, problems need to be solved in the existing methods, such as detection accuracy and interference of complex background textures. In this paper, we propose an efficient segmentation algorithm that combines traditional operators with deep learning networks to alleviate the existing problems. Specifically, we introduce a Gabor filter into the model, which provides the unique advantage of extracting low-level texture features to solve the problem of texture interference and enable the algorithm to converge quickly in the early stages of training. Furthermore, we design a U-shaped architecture that is not completely symmetrical, making model training easier. Meanwhile, multi-stage result fusion is proposed for precise location of defects. The design of this framework significantly improves the detection accuracy and effectively breaks through the limitations of transformer-based models. Experimental results show that on a dataset with one class, a small amount of data, and complex sample background texture, our method achieved 90.03% and 33.70% in ACC and IoU, respectively, which is almost 10% higher than other previous state of the art models. Experimental results based on three different fabric datasets consistently show that the proposed model has excellent performance and great application potential in the industrial field.

Embedding Global Contrastive and Local Location in Self-Supervised Learning

Self-supervised representation learning (SSL) typically suffers from inadequate data utilization and feature-specificity due to the suboptimal sampling strategy and the monotonous optimization method. Existing contrastive-based methods alleviate these issues through exceedingly long training time and large batch size, resulting in non-negligible computational consumption and memory usage. In this paper, we present an efficient self-supervised framework, called GLNet. The key insights of this work are the novel sampling and ensemble learning strategies embedded in the self-supervised framework. We first propose a location-based sampling strategy to integrate the complementary advantages of semantic and spatial characteristics. Whereafter, a Siamese network with momentum update is introduced to generate representative vectors, which are used to optimize the feature extractor. Finally, we particularly embed global contrastive and local location tasks in the framework, which aims to leverage the complementarity between the high-level semantic features and low-level texture features. Such complementarity is significant for mitigating the feature-specificity and improving the generalizability, thus effectively improving the performance of downstream tasks. Extensive experiments on representative benchmark datasets demonstrate that GLNet performs favorably against the state-of-the-art SSL methods. Specifically, GLNet improves MoCo-v3 by 2.4% accuracy on ImageNet dataset, while improves 2% accuracy and consumes only 75% training time on the ImageNet-100 dataset. In addition, GLNet is appealing in its compatibility with popular SSL frameworks. Code is available at GLNet.

RGB-T Semantic Segmentation With Location, Activation, and Sharpening

Semantic segmentation is important for scene understanding. To address the scenes of adverse illumination conditions of natural images, thermal infrared (TIR) images are introduced. Most existing RGB-T semantic segmentation methods follow three cross-modal fusion paradigms, i.e. encoder fusion, decoder fusion, and feature fusion. Some methods, unfortunately, ignore the properties of RGB and TIR features or the properties of features at different levels. In this paper, we propose a novel feature fusion-based network for RGB-T semantic segmentation, named \emph{LASNet}, which follows three steps of location, activation, and sharpening. The highlight of LASNet is that we fully consider the characteristics of cross-modal features at different levels, and accordingly propose three specific modules for better segmentation. Concretely, we propose a Collaborative Location Module (CLM) for high-level semantic features, aiming to locate all potential objects. We propose a Complementary Activation Module for middle-level features, aiming to activate exact regions of different objects. We propose an Edge Sharpening Module (ESM) for low-level texture features, aiming to sharpen the edges of objects. Furthermore, in the training phase, we attach a location supervision and an edge supervision after CLM and ESM, respectively, and impose two semantic supervisions in the decoder part to facilitate network convergence. Experimental results on two public datasets demonstrate that the superiority of our LASNet over relevant state-of-the-art methods. The code and results of our method are available at https://github.com/MathLee/LASNet.

Statistical Texture Learning Method for Monitoring Abandoned Suburban Cropland Based on High-Resolution Remote Sensing and Deep Learning

Cropland abandonment is crucial in agricultural management and has a profound impact on crop yield and food security. In recent years, many cropland abandonment identification methods based on remote sensing observation data have been proposed, but most of these methods are based on coarse-resolution images and use traditional machine learning methods for simple identification. To this end, we perform abandonment recognition on high-resolution remote sensing images. According to the texture features of the abandoned land, we combine the method of statistical texture learning and propose a new deep learning framework called PSPNet-STL. The model integrates high-level semantic feature extraction and deep mining of low-level texture features to identify cropland abandonment. First, we labeled the abandoned cropland area and built the HRAC dataset, a high-resolution cropland abandonment dataset. Secondly, we improved PSPNet by fusing statistical texture learning modules to learn multiple texture information on low-level feature maps, and combined high-level semantic features for cropland abandonment recognition. Experiments are performed on the HRAC dataset. Compared with other methods, the proposed model has the best performance on this dataset, both in terms of accuracy and visualization, proving that deep mining of low-level statistical texture features is beneficial for crop abandonment recognition.

A Novel Heterogeneous Network for Modeling Driver Attention With Multi-Level Visual Content

Driver attention modeling is a crucial technique in building human-centric intelligent driving systems. Considering the human visual mechanism, this study leverages multi-level visual content, including low-level texture features, middle-level optical flows, and high-level semantic information, as the model input. Subsequently, a heterogeneous model is proposed to handle the multi-level input, which integrates the graph and convolutional neural networks. Distinguished from the existing studies that use semantic segmentation, our study directly leverages the objection detection information in an interpretable manner. To deal with the detected objects, in this work, a graph attention network is used to explicitly construct the semantic information, rather than handle the features extracted by convolutional modules for building the latent space features, which are used in existing studies. Further, a semantic attention module is proposed to integrate the non-Euclidean output of the graph network with the Euclidean feature maps of the convolutional neural networks. Finally, these integrated features are decoded to generate a driver attention map. Three typical datasets are used to validate the proposed method. A comprehensive comparison and analysis have proven the feasibility and validity of our proposed method, as well as its ability to achieve state-of-the-art performance.

Cell Phenotype Classification Based on Joint of Texture Information and Multilayer Feature Extraction in DenseNet.

Cell phenotype classification is a critical task in many medical applications, such as protein localization, gene effect identification, and cancer diagnosis in some types. Fluorescence imaging is the most efficient tool to analyze the biological characteristics of cells. So cell phenotype classification in fluorescence microscopy images has received increased attention from scientists in the last decade. The visible structures of cells are usually different in terms of shape, texture, relationship between intensities, etc. In this scope, most of the presented approaches use one type or joint of low-level and high-level features. In this paper, a new approach is proposed based on a combination of low-level and high-level features. An improved version of local quinary patterns is used to extract low-level texture features. Also, an innovative multilayer deep feature extraction method is performed to extract high-level features from DenseNet. In this respect, an output feature map of dense blocks is entered in a separate way to pooling and flatten layers, and finally, feature vectors are concatenated. The performance of the proposed approach is evaluated on the benchmark dataset 2D-HeLa in terms of accuracy. Also, the proposed approach is compared with state-of-the-art methods in terms of classification accuracy. Comparison of results demonstrates higher performance of the proposed approach in comparison with some efficient methods.

Exploiting deep and hand-crafted features for texture image retrieval using class membership

• Proposed a novel scheme for texture image retrieval. • Proposed scheme exploits the strength of both hand-crafted and deep features. • Modified distance function is used for image retrieval in wavelet feature space. • Distance function is weighted using output class memberships of the query image. • Efficiency of the proposed method found to be promising. In the modern digital era, with the availability of low-cost hardware like sensors and cameras, a huge amount of image databases are being created for diverse applications. These databases give rise to the need of developing efficient content-based image retrieval (CBIR) systems. Major efforts have been put over the past two decades to develop different global and low-level texture features to build efficient CBIR systems. However, designing texture features that are suitable for distance-based retrieval is always a challenging task. Recently, Convolution Neural Networks have shown promising results for object detection and classification . CNNs are also applied to build classifier-based retrieval systems. However, the classifier-based retrieval methods can retrieve images only from the predicted class. Therefore, the performance of such system greatly depends on classification performance of the classifier . This paper proposes a method that exploits the strength of the Convolutional Neural Networks for predicting the class membership of the query image for all output classes and retrieve images using a modified distance function in the wavelet feature space. The performance of the proposed method is evaluated using three popular texture datasets of varying complexity and found to be superior to all competing methods considered.

LLAM-MDCNet for Detecting Remote Sensing Images of Dead Tree Clusters

Clusters of dead trees are forest fires-prone. To maintain ecological balance and realize its protection, timely detection of dead trees in forest remote sensing images using existing computer vision methods is of great significance. Remote sensing images captured by Unmanned aerial vehicles (UAVs) typically have several issues, e.g., mixed distribution of adjacent but different tree classes, interference of redundant information, and high differences in scales of dead tree clusters, making the detection of dead tree clusters much more challenging. Therefore, based on the Multipath dense composite network (MDCN), an object detection method called LLAM-MDCNet is proposed in this paper. First, a feature extraction network called Multipath dense composite network is designed. The network’s multipath structure can substantially increase the extraction of underlying and semantic features to enhance its extraction capability for rich-information regions. Following that, in the row, column, and diagonal directions, the Longitude Latitude Attention Mechanism (LLAM) is presented and incorporated into the feature extraction network. The multi-directional LLAM facilitates the suppression of irrelevant and redundant information and improves the representation of high-level semantic feature information. Lastly, an AugFPN is employed for down-sampling, yielding a more comprehensive representation of image features with the combination of low-level texture features and high-level semantic information. Consequently, the network’s detection effect for dead tree cluster targets with high-scale differences is improved. Furthermore, we make the collected high-quality aerial dead tree cluster dataset containing 19,517 images shot by drones publicly available for other researchers to improve the work in this paper. Our proposed method achieved 87.25% mAP with an FPS of 66 on our dataset, demonstrating the effectiveness of the LLAM-MDCNet for detecting dead tree cluster targets in forest remote sensing images.

A Cascaded Twin Dense Network Based Metallic Surface Defect Detection Algorithm

A cascaded twin dense network based metallic surface defect detection algorithm SCSEG-Net is proposed to address the problems of low accuracy and difficulty in locating defects in real-time metal surface defect detection. The method incorporates atrous spatial pyramid pooling structure to obtain feature maps with different sampling rates to capture multi-scale information. Meanwhile, in order to enhance the classification accuracy, the low-level texture and boundary features obtained by shallow convolution and the complex high-level feature information obtained by deep convolution are fused during training to better optimize the training parameters through the cascade network. SCSEG-Net can convert defect images into pixel-level prediction masks and quickly acquire real defect classes. The SCSEG-Net algorithm is trained, evaluated and verified on the industry standard steel surface defect data set. The results show that compared with similar algorithms, proposed method can segment the contour of steel surface defects more accurately and complete the classification with an <italic>F</italic>₁-score of 97.8% and a recall rate of 98.81%.

Completed local binary patterns feature integrated convolutional neural network-based terrain classification algorithm in polarimetric synthetic aperture radar images

We propose a polarimetric synthetic aperture radar (PolSAR) image terrain classification algorithm based on complete local binary patterns (CLBP) feature integrated convolutional neural network (CNN) (CLBP-CNN). Traditional CNN has a powerful high-level deep features extraction ability, which can effectively improve the terrain classification accuracy in PolSAR images. However, most traditional CNN-based methods only focus on the high-level deep feature extraction of the synthetic aperture radar (SAR) terrains; they ignore the low-level texture features, resulting in incomplete feature representation and poor classification accuracy. In fact, low-level texture features also play an important role in PolSAR terrain classification. To solve the problem that traditional CNN-based terrain classification methods easily lose the low-level texture features in the process of feature extraction, the proposed method uses the CLBP descriptor to extract multi-level texture features under different receptive fields, and it adaptively combines the high-level deep features and the low-level texture features for better SAR terrain feature description. CLBP-CNN greatly alleviates the shortcomings of traditional CNN in missing the low-level texture features; it improves the feature representation completeness, so it can achieve better terrain classification results. The superiority of CLBP-CNN is verified on the data sets of Flevoland, San Francisco, and Oberpfaffenhofen.

A Cascaded Feature Extraction for Diagnosis of Ovarian Cancer in CT Images

This paper proposed ovarian cancer detection in the ovarian image using joint feature extraction and an efficient Net model. The noise of the input image is filtered by using Improved NLM (Improved Non-Local Means) filtering. The deep features are extracted using Deep CNN_RSO (Deep Convolutional Neural Network Rat Swarm Optimization), and the low-level texture features are extracted using ILBP (Interpolated Local Binary Pattern or Interpolated LBP). To improve the feature extraction and reduce the error, use a cascading technique for the feature extraction. RSO also helps to efficiently optimize the DCNN features from the images. Finally, the extracted image is classified using the Efficient Net classifier, which performs a global average summary and classification of ovarian cancer (normal and abnormal). The system’s performance is implemented on the Cancer Genome Atlas Ovarian Cancer (TCGA-OV) dataset. The system’s performance, like sensitivity, specificity, accuracy and error rates, shows better with respect to other techniques.

Dense Haze Removal Based on Dynamic Collaborative Inference Learning for Remote Sensing Images

Haze in remote sensing images (RSIs) usually causes serious radiance distortion and image quality degeneration, resulting in difficult remote sensing inversion and interpretation. Under the condition of dense haze, existing dehazing methods still experience problems to be solved: 1) the texture details and spectral characteristics in RSIs cannot be restored well; 2) small-scale objects, such as cars and ships, which often consist of only a few pixels in RSIs, cannot be effectively highlighted in dehazed results. To solve these issues, we propose a novel dynamic collaborative inference learning (DCIL) framework that can significantly restore real surface information from dense hazy RSIs. First, we design a dynamic mutual enhancement (DME) mechanism to reinforce the low-level texture features by integrating primary information and semantic information at different levels. Second, we propose a spectrum-aware aggregation (SAA) strategy to mine the spectrum features among multiscale restored results, which can fully capture spectral characteristics. Third, we build a collaborative criterion by constructing a Siamese network structure in the training stage to improve the robustness and generalization performance of DCIL considering the diversity of the scale range and view change of RSIs. Finally, we propose a phased learning strategy to deduce the implicit haze-relevant features by gradually increasing the concentration of haze which can effectively address small-scale objects obscured by dense haze. To this end, we develop two synthetic remote sensing dehazing datasets to train our model, which can also alleviate the dilemma of hazy RSI datasets shortages. Experimental results on both synthetic datasets and real remote sensing hazy images demonstrate that the proposed DCIL can attain significant progress compared to competing methods. The two synthetic hazy datasets are available at https://github.com/Shan-rs/DCI-Net.

A Joint Optimization Framework of the Embedding Model and Classifier for Meta-Learning

The aim of meta-learning is to train the machine to learn quickly and accurately. Improving the performance of the meta-learning model is important in solving the problem of small samples and in achieving general artificial intelligence. A meta-learning method based on feature embedding that exhibits good performance on the few-shot problem was previously proposed. In this method, the pretrained deep convolution neural network was used as the embedding model of sample features, and the output of one layer was used as the feature representation of samples. The main limitation of the method is the inability to fuse low-level texture features and high-level semantic features of the embedding model and joint optimization of the embedding model and classifier. Therefore, a multilayer adaptive joint training and optimization method of the embedding model was proposed in the current study. The main characteristics of the current method include using multilayer adaptive hierarchical loss to train the embedding model and using the quantum genetic algorithm to jointly optimize the embedding model and classifier. Validation was performed based on multiple public datasets for meta-learning model testing. The proposed method shows higher accuracy compared with multiple baseline methods.

EDDs: A series of Efficient Defect Detectors for fabric quality inspection

Deep Convolutional Neural Network (DCNN) has recently advanced state-of-the-art performance on vision-related tasks and its application is further extended to industrial fields. The paper focuses on the problem of fabric defect detection to which an efficient DCNN architecture is developed. In contrast to previous methods that directly apply existing DCNN models demonstrated on natural images to industrial images, the proposed Efficient Defect Detectors (EDDs) are sufficiently optimized with consideration of the characteristics of fabric surface images, i.e., resolution, defect appearance, etc. Firstly, lightweight backbone is suggested in EDD to improve computational efficiency without reduction in image resolution. Secondly, a new feature fusion strategy named L-shaped feature pyramid network (L-FPN) is proposed and utilized to make full use of low-level texture features which are demonstrated to be more important than high-level semantic features in defect recognition. Based on the configurations of lightweight backbone and L-FPN, we use only one hyper-parameter to jointly adjust the proportion of resources occupied by width, depth and input resolution so that a family of defect detectors under different resource constraints can be developed. Experiments are conducted on a large fabric dataset to demonstrated the effectiveness of EDDs. Compared with the recent state-of-the-art detector, EfficientDet-d3, EDD-d3 achieves higher mean Average Precision (mAP) (20.9 vs 19.9) but with fewer parameters. EDD-d3 has 8.59M parameters and 31.78B FLOPs (floating point operation per second), which respectively are 39.8% and 49.0% lower than EfficientDet-d3. The proposed EDDs achieve better trade-off between accuracy and speed than previous methods. EDDs could be applied to fabric production sits with different resource restrictions, which demonstrates that EDDs have important application value.

Saliency detection using a deep conditional random field network

Saliency detection has made remarkable progress along with the development of deep learning. While how to integrate the low-level intrinsic context with high-level semantic information to keep the object boundary sharp and restrain the background noise is still a challenging problem. Many attempts on network structures and refinement strategies have been explored, such as using Conditional Random Field (CRF) to improve the accuracy of saliency map, but it is independent from the deep network and cannot be trained end-to-end. To tackle this issue, we propose a novel Deep Conditional Random Field network (DCRF) to take both deep feature and neighbor information into consideration. First, Multi-scale Feature Extraction Module (MFEM) is adopted to capture the low level texture and high level semantic features, multi-stacks of deconvolution layers are employed to improve the spatial resolution of deep layers. Then we employ Backward Optimization Module (BOM) to guide shallower layers by high-level location and shape information derived from deeper layers, which intrinsically enhance the representational capacity of low-level features. Finally, a Deep Conditional Random Field Module (DCRFM) with unary and pairwise potentials is designed to concentrate on spatial neighbor relations to obtain a compact and uniformed saliency map. Extensive experimental results on 5 datasets in terms of 6 evaluation metrics demonstrate that the proposed method achieves state-of-the-art performance.

Multi-Scale Rotation-Invariant Haar-Like Feature Integrated CNN-Based Ship Detection Algorithm of Multiple-Target Environment in SAR Imagery

This paper proposes a multi-scale rotation-invariant haar-like (MSRI-HL) feature integrated convolutional neural network (MSRIHL-CNN)-based ship detection algorithm of the multiple-target environment in synthetic aperture radar (SAR) imagery. Usually, ship detection includes preprocessing, prescreening, discrimination, and classification. Among them, prescreening and discrimination are the most two important stages so that they catch great intention. Based on our previous work, we propose a truncated-clutter-statistics-based joint, constant false alarm rate (CFAR) detector (TCS-JCFAR) for ship target prescreening in the multiple-target environment. TCS-JCFAR greatly enhances the prescreening rate in the multiple-target environment while achieving a low observed FAR. In the discrimination stage, conventional CNN extracts the deep features (high-level features); however, it will lose the local texture and edge information (low-level features) which are of great significance for target discrimination. Hence, the MSRI-HL features are used to represent the multi-scale, rotation-invariant texture, and edge information that conventional CNN fails to capture. The extracted low-level MSRI-HL features and the high-level deep features are optimally fused to a multi-layered feature vector. Finally, the multi-layered feature vector is fed into a typical support vector machine (SVM) classifier for ship target discrimination. The proposed MSRIHL-CNN combines the low-level texture and edge features and the high-level deep features; moreover, they are optimally fused to fully represent the ship targets. Undoubtedly, MSRIHL-CNN has better discrimination performance. The superiority of the proposed TCS-JCFAR-based prescreener and MSRIHL-CNN-based discriminator is validated on the Chinese Gaofen-3 SAR imagery.

Improved Deep Learning Network Based in combination with Cost-sensitive Learning for Early Detection of Ovarian Cancer in Color Ultrasound Detecting System.

With the development of theories and technologies in medical imaging, most of the tumors can be detected in the early stage. However, the nature of ovarian cysts lacks accurate judgement, leading to that many patients with benign nodules still need Fine Needle Aspiration (FNA) biopsies or surgeries, increasing the physical pain and mental pressure of patients as well as unnecessary medical health care costs. Therefore, we present an image diagnosis system for classifying the ovarian cysts in color ultrasound images, which novelly applies the image features fused by both high-level features from deep learning network and low-level features from texture descriptor. Firstly, the ultrasound images are enhanced to improve the quality of training data set and the rotation invariant uniform local binary pattern (ULBP) features are extracted from each of the images as the low-level texture features. Then the high-level deep features extracted by the fine-tuned GoogLeNet neural network and the low-level ULBP features are normalized and cascaded as one fusion feature that can represent both the semantic context and the texture patterns distributed in the image. Finally, the fusion features are input to the Cost-sensitive Random Forest classifier to classify the images into "malignant" and "benign". The high-level features extracted by the deep neural network from the medical ultrasound image can reflect the visual features of the lesion region, while the low-level texture features can describe the edges, direction and distribution of intensities. Experimental results indicate that the combination of the two types of features can describe the differences between the lesion regions and other regions, and the differences between lesions regions of malignant and benign ovarian cysts.

Automatic Example-based Image Colourisation using Location-Aware Cross-Scale Matching.

Given a reference colour image and a destination grayscale image, this paper presents a novel automatic colourisation algorithm that transfers colour information from the reference image to the destination image. Since the reference and destination images may contain content at different or even varying scales (due to changes of distance between objects and the camera), existing texture matching based methods can often perform poorly. We propose a novel cross-scale texture matching method to improve the robustness and quality of the colourisation results. Suitable matching scales are considered locally, which are then fused using global optimisation that minimises both the matching errors and spatial change of scales. The minimisation is efficiently solved using a multi-label graph-cut algorithm. Since only low-level texture features are used, texture matching based colourisation can still produce semantically incorrect results, such as meadow appearing above the sky. We consider a class of semantic violation where the statistics of up-down relationships learnt from the reference image are violated and propose an effective method to identify and correct unreasonable colourisation. Finally, a novel nonlocal ℓ1 optimisation framework is developed to propagate high confidence micro-scribbles to regions of lower confidence to produce a fully colourised image. Qualitative and quantitative evaluations show that our method outperforms several state-of-the-art methods.