Efficient local and global contour detection based on superpixels

The object contour detection in natural scenes is an important research problem in computer vision. The difficulty is that the integrity of contour extracted is interfered with a large number of texture edges in the background seriously. In recent years, some researchers have introduced image contour detection method with biological visual features and achieved preferable results. Such as, based on the fact of visual outer region suppression, they can suppress a certain amount of texture edge while extracting the contour of the object, and obtain the proper contour edge. However, when this method is used for contour detection in some image which the texture is similar to the contour, the texture edge still cannot be removed well, and the result is unsatisfactory. In consideration of the merit and demerit of the feature suppressed method, we propose an improved algorithm for object contour detection in natural scene based on image saliency. Firstly, the initial contour is obtained by using the suppression effect of non-classical receptive field, and the saliency map is obtained by using the hyper-complex Fourier transform image saliency detection method, and then the initial contour and the saliency map are integrated into the final contour. Compared with the traditional contour detection method, the contour extraction algorithm based on image saliency can remove more background information and improve the accuracy and integrity of object contour extracted in natural scenes.

An approach to symbolic contour extraction is described that consists of three stages: enhancement, detection, and extraction of contours and corners. Contours and corners are enhanced by models of monkey cortical complex and endstopped cells. Detection of corners and local contour maxima is performed by selection of local maxima in both contour and corner enhanced images. These maxima form the anchor points of a greedy contour following algorithm that extracts the contours. This algorithm is based on the idea of spatially linking neurons along a contour that fire in synchrony to indicate an extracted contour. The extracted contours and detected corners represent the symbolic representation of the image. The advantage of the proposed model over other models is that the same low constant thresholds for corner and local contour maxima detection are used for different images. Closed contours are guaranteed by the contour following algorithm to yield a fully symbolic representation which is more suitable for reasoning and recognition. In this respect our methodology is unique, and clearly different from the standard (edge) contour detection methods. The results of the extracted contours (when displayed as being detected) show similar or better results compared to the SUSAN and Canny-CSS detectors.

The complex hairiness of slub yarns makes it challenging to accurately detect the yarn body and measure its appearance parameters. This paper proposes a method for measuring the parameters of continuous slub yarn images based on computer vision. First, overlapping slub yarn sequence images are captured using a specially designed image capture device. Then, an image stitching method based on normalized cross-correlation is proposed to remove the overlapping data from neighboring slub yarn images. To handle the complex hairiness around the slub yarn, an improved knowledge-enhanced contour detection method is used to accurately extract the backbone contour of the slub yarn. Finally, the extracted slub yarn contour is analyzed using such methods as frequency histograms and curve fitting. Contour detection results show that the proposed yarn contour detection method achieves better performance, with an optimal dataset scale of 0.930 and an optimal image scale of 0.931. The parameters of slub yarn measured using the proposed method are consistent with the production process settings. The experimental results demonstrate that the method can accurately detect slub yarn parameters, providing a basis for the appearance design and proofing of slub fabrics in textile enterprises.

It is an important task to count the number of the objects in the image automatically. Because the objects in the image are sometimes overlapped with each other and even covered with different shaped objects, it is very difficult to detach these objects before counting them. Circular Hough transform has been used extensively to detect and count overlapped objects with circular shapes. However it can not detach the overlapping circular objects efficiently, and the counting task will provide inaccurate results when the circular objects are covered with random shaped objects. To deal with this problem, an integrated method is proposed based on the Hough Transform and Contour Detection methods. Furthermore, a software is implemented using the open source computer vision library Opencv 3.0 with visual studio 2012. The experimental results show that the proposed method can detach and count the objects accurately.

This paper investigates two fundamental problems in computer vision: contour detection and image segmentation. We present state-of-the-art algorithms for both of these tasks. Our contour detector combines multiple local cues into a globalization framework based on spectral clustering. Our segmentation algorithm consists of generic machinery for transforming the output of any contour detector into a hierarchical region tree. In this manner, we reduce the problem of image segmentation to that of contour detection. Extensive experimental evaluation demonstrates that both our contour detection and segmentation methods significantly outperform competing algorithms. The automatically generated hierarchical segmentations can be interactively refined by user-specified annotations. Computation at multiple image resolutions provides a means of coupling our system to recognition applications.

Background Recently, high temporal resolution short-axis cine images have been used to create left ventricular (LV) regional mechanical delay (RMD) maps in patients undergoing cardiac resynchronization therapy (CRT). Generating the endocardial segmentation required for RMD analysis is both time-consuming and subjective. A semi-automated segmentation method for time-continuous contour detection that uses multi-dimensional dynamic programming to enforce user-defined shape constraints without the need for extensive training sets has been previously developed (Uzumcu 2006). Our goal was to compare RMD maps generated from semiautomated and manual segmentation in patients enrolled for CRT.

This research focuses on the development of a real-time vehicle detection and tracking system using OpenCV and deep learning models to address growing traffic management challenges. The project integrates the Gaussian Mixture-based Background Subtraction (MOG2), morphological operations, contour detection, and tracking methods like Kalman Filtering and Deep SORT. Additionally, object classification via YOLO and MobileNet-SSD significantly improves the accuracy of identifying and distinguishing vehicles. Evaluated on multiple real-world datasets under various environmental conditions, the system demonstrated strong performance with over 91% mAP, 92.5% precision, and robust multi-object tracking accuracy. This study serves as a scalable framework for intelligent traffic surveillance.

X-ray angiography is a common method for image-guided navigation comprising surgical devices and vessel contours detection which can offer visual feedback to physicians. However, for the sake of both physicians' and patients' health, the contrast agent is injected intermittently and in a low dosage, so the X-ray images are not always of adequate quality for visual examination. So the navigation, which is to achieve the real-time positions of vessels and devices, is a challenging task. On one hand, the vessels and devices are always affected by other tissues. On the other hand, when the contrast agent is absent, vessels are invisible, so it is impossible to detect vessel contours. Aimed at the above problems, this paper proposes a vessel detection and estimation method. The paper has two main contributions. First, the paper uses the multiple scales top-hat enhancement method to improve image contrast rate, which is benefit for the vessel contours detection. Second, the paper proposes a method to estimate the vessel contours when the contrast agent is absent, it is based on the fact that vessel moves in a small range and the vessel contours are estimated by using the contour detection results of the adjacent frames. The experiment has been tested on 18 sequences images and the average accuracy is 92.6%. The subjective evaluation and the experiment results demonstrate the effectiveness and robust of the proposed method.

PurposeTo overcome the severe intensity inhomogeneity and blurry boundaries in HIFU (High Intensity Focused Ultrasound) ultrasound images, an accurate and efficient multi-scale and shape constrained localized region-based active contour model (MSLCV), was developed to accurately and efficiently segment the target region in HIFU ultrasound images of uterine fibroids.MethodsWe incorporated a new shape constraint into the localized region-based active contour, which constrained the active contour to obtain the desired, accurate segmentation, avoiding boundary leakage and excessive contraction. Localized region-based active contour modeling is suitable for ultrasound images, but it still cannot acquire satisfactory segmentation for HIFU ultrasound images of uterine fibroids. We improved the localized region-based active contour model by incorporating a shape constraint into region-based level set framework to increase segmentation accuracy. Some improvement measures were proposed to overcome the sensitivity of initialization, and a multi-scale segmentation method was proposed to improve segmentation efficiency. We also designed an adaptive localizing radius size selection function to acquire better segmentation results.ResultsExperimental results demonstrated that the MSLCV model was significantly more accurate and efficient than conventional methods. The MSLCV model has been quantitatively validated via experiments, obtaining an average of 0.94 for the DSC (Dice similarity coefficient) and 25.16 for the MSSD (mean sum of square distance). Moreover, by using the multi-scale segmentation method, the MSLCV model’s average segmentation time was decreased to approximately 1/8 that of the localized region-based active contour model (the LCV model).ConclusionsAn accurate and efficient multi-scale and shape constrained localized region-based active contour model was designed for the semi-automatic segmentation of uterine fibroid ultrasound (UFUS) images in HIFU therapy. Compared with other methods, it provided more accurate and more efficient segmentation results that are very close to those obtained from manual segmentation by a specialist.

Recently, most of the people working in Medical field need a CAD system to investigate and analyze the diseases to provide appropriate treatment and prevent from diseases in the earlier stage itself In this paper an automatic Optic Disc detection for analyzing and diagnosing the disease level in the fundus images has been introduced The retina has the features like optic disc, macula, fovea, micro aneurysms, exudates. The optic disc is detected and removed, since it has the homogeneity with the abnormal feature like exudates. This CAD system helps the ophthalmologist to do further investigation and prior knowledge about the disease. The proposed approach contains a sequence of steps where the fundus image has been preprocessed by histogram equalization and the optic disc is detected by using background subtraction and contour detection method.

Optical Character Recognition is one of the challenging areas in pattern recognition systems. It has gained a lot of importance in the fields of image processing and computer vision. In this paper, the handwritten digits on an evaluated answer booklet are taken for recognition by the neural network. Image acquisition, connected component analysis and contour detection methods are performed prior to the recognition process. The segmented handwritten digits are recognized through neural networks based on Auto Resonance Network as described in this paper. Auto Resonance Network can classify real valued multi-dimensional inputs and can have an adjustable threshold for each node in the network. The Modified National Institute of Standards and Technology database is used for training the network. An accuracy of 91% is achieved by using a single layer Auto Resonance Network.

In this Paper, the human body parts are measured using pose estimation method. Contour detection method is applied to detect the edges. This work aims to detect the body pose and track the sematic key points, which helps to measure the distances between various body parts. Media pipe model is used in this work, which is highly accurate for tracing sematic key-points in human body parts estimation. Edge detection method and CV2 are used to find the boundaries of the human body. In brief, this paper aims to provide a simple approach that will serve as a solid baseline for body parts measurement.

This paper describes a method for automatic contour detection in long-axis cardiac MRI using an adaptive virtual exploring robot. The robot is a simulated trained virtual autonomous tri-cycle that is initially positioned in a binary representation of the left ventricle (LV) and finds the contours during navigation through the ventricle. The method incorporates global and local prior shape knowledge of the LV in order to adapt the navigational parameters. Together with kinematic constraints, the robot is able to avoid concave regions such as papillary muscles and navigate through narrow corridors such as the apex. Validation was performed on in-vivo multiphase long-axis cardiac MRI images of 11 subjects. Results showed good correlation between the quantitative parameters, computed from manual and automatic segmentation: for end-diastolic volume (EDV) r=0.91, for end-systolic volume (ESV) r=0.93, ejection fraction (EF) r=0.77, and LV mass (LVM) r=0.80.

Contour detection is an important step in contour based object analysis in 2D images. We present a novel method for contour detection optimized for contour traversal tasks in 2D images. The method is based on morphological operations and detects a one-pixel wide contour around objects in binary images. Performance of the method is evaluated on ground truth images from our dataset created by manually tracing the contour in images. Proposed method achieves high DICE coefficient overlap to ground truth images and superior detection compared to other state-of-the-art methods.

We apply a cellular neural network (CNN) based contour detection to exact the main reflective interface in seismic image. The seismic image is generated by reverse time migration (RTM) using the pseudospectral time domain (PSTD) method. According to Nyquist sampling theorem, the PSTD algorithm requires only two points per minimum wavelength rather than the traditional high order finite difference time domain (FDTD) which needs more than eight points per minimum wavelength to get the same accuracy. Thus, RTM using the PSTD algorithm can reduce the computing costs greatly when comparing with the traditional RTM based on the FDTD algorithm. To get more clear reflective interface in imaging result of reverse time migration, we use a contour detection algorithm based on CNN which calculates the template parameters by the gray-scale and spatial relationship between the central pixel and the other neighboring pixels in the current local window. The simulation results shows that the proposed method have good efficiency and imaging quality, and the contour detection method makes the reflective interface easier to distinguish.