Background Subtraction Using Local SVD Binary Pattern

Guo, Lili; Xu, Dan; Qiang, Zhenping

doi:10.1109/cvprw.2016.148

Cited by 48 publications

(40 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…approaches, are all quite mature and robust to varying lighting conditions, clothing and even substantial occlusion. On the other hand, leveraging the fact that cameras are stationary, Background Subtraction techniques, such as Guo et al (2016) can be used to "hide" all the products that remain on the shelf and therefore focus the attention of the visual classifier only on the moving foreground, where it can find the item being taken or returned.…”

Section: Vision Event Pre-processingmentioning

confidence: 99%

FAIM: Vision and Weight Sensing Fusion Framework for Autonomous Inventory Monitoring in Convenience Stores

Falcão

Ruiz²,

Pan

et al. 2020

Front. Built Environ.

View full text Add to dashboard Cite

A common pain point for physical retail stores is live inventory monitoring, i.e., knowing how many items of each product are left on the shelves. About 4% of sales are lost due to an average 5-10% out-of-shelf stockout rate, while additional supplies existed in the warehouse. Traditional techniques rely on manual inspection, per-item tagging using RFIDs, or human-in-the-loop systems, such as Amazon Go. These approaches, while effective, either have poor accuracy, long delays between results or are cost prohibitive. In this paper, we present FAIM (Autonomous Inventory Monitoring Framework) for cashier-less stores. To the best of our knowledge, this is the first fully autonomous system that fuses multiple sensing modalities. Utilizing weight difference on a shelf, visual item recognition in customers' hands and prior knowledge of item layout FAIM monitors products picked up or returned without human-in-the-loop. We present results from a real-world setup with 85 items (33 unique products) replicating the layout of a local 7-Eleven store. To evaluate our system we characterize the similarity of the unique products across three physical features (i.e., weight, color, and location). Our results show that the fused approach provides up to 92.6% item identification accuracy, a 2× reduction in error compared to reported self-checkout stations.

show abstract

Section: Vision Event Pre-processingmentioning

confidence: 99%

FAIM: Vision and Weight Sensing Fusion Framework for Autonomous Inventory Monitoring in Convenience Stores

Falcão

Ruiz²,

Pan

et al. 2020

Front. Built Environ.

View full text Add to dashboard Cite

show abstract

“…This does not include foreground/background segmentation (FG/BG), as this step is already very well covered by o-the-shelf components of OpenCV. 13…”

Section: Software Implementationmentioning

confidence: 99%

Data fusion of multi-spectral cameras on a low-power processing platform for self-sufficient outdoor operation

Reichel¹,

Peter²,

Döge³

et al. 2019

Photonics and Education in Measurement Science 2019

View full text Add to dashboard Cite

Multi-spectral camera set-ups may generally allow for creating surveillance applications even under unfavorable conditions, such as low-light environments or scenes involving vastly dierent lighting conditions. A highresolution color camera, a high-dynamic-range camera and an infrared thermal camera were combined into a self-sucient platform for continuous outdoor operation. The sheer amount of produced data poses a serious challenge, both in terms of available bandwidth and processing power, because self-suciency requires using relatively low-power components, and privacy, as high-resolution, multi-spectral image data are sensitive information. Thus, relevant objects of interest had to be eciently extracted, tracked and georeferenced on the sensor platform. These data, from one or more sensorheads, are then sent via WLAN or mobile data link to a central control unit, possibly anonymized, e.g. prompting immediate action by a human operator in a disaster response use case, or stored for further oine analysis when used in the framework of Smart City. Applying the classic stereo vision approach would require calibrating both intrinsic and extrinsic parameters of all cameras. The input data's multi-spectral nature complicates the correspondence problem for extrinsic parameter calibration and subsequent stereo matching, while intrinsic parameter calibration according to the pinhole camera model is made dicult due to the cameras having to be focused at innity. However, by making certain reasonable assumptions about the observed scene in typical use cases, accepting a possible loss in localization accuracy, camera calibration could be limited to the bare minimum and less computational power was required at run-time.

show abstract

“…The LBP technique has been used since its proposition as a powerful resource for texture classification. The technique has also been widely used in facial detection and unsupervised texture segmentation [14], [15]. A variation of LBP was also used to calculate the texture contrast of the pixels of an image.…”

Section: B Preprocessingmentioning

confidence: 99%

“…Among the different image processing techniques, Local Binary Patterns (LBP) [13] features a robust form of texture classification. After its first publication, the technique has undergone several improvements, LBP modifications have been published, and the technique can currently be used for face detection and unsupervised texture segmentation [14], [15].…”

Section: Introductionmentioning

confidence: 99%

Image Processing Techniques to Improve Deep 6DoF Detection in RGB Images

Felix¹,

Simões

Cunha³

et al. 2019

Anais Estendidos Do Simpósio De Realidade Virtual E Aumentada (SVR)

View full text Add to dashboard Cite

Six degrees of freedom (6DoF) Object Detection has great relevance in computer vision due to its use in applications on several areas, such as augmented reality and robotics. Even with the improved results provided by deep learning techniques, object detection of textured and non-textured objects is still a challenge. The objective of this work was to seek improvements in the six degrees of freedom detection of non-textured objects using a Convolutional Neural Network (CNN) approach through the preprocessing of the images that were used for training the network. A State of the art research was carried out on techniques that use CNN to detect objects in six degrees of freedom. We also searched for filters with enhancement factors for detection. Finally, a detection technique based on a CNN was selected and adapted to use single-channel images (grayscale) as input, instead of using three-channel images (RGB) as in the original proposition, aiming to increase its robustness while reducing the complexity of the input images. The technique was also tested with the application of two different preprocessing filters to enhance the objects’ contours on the single-channel images, one being the ”pencil effect”, and the other based on local binary patterns (LBP). With this study, it was possible to evaluate the impact on the CNN detection performance due to the application of both of the filters. The proposed technique used with one channel images and the filters on the images still could not surpass the results of the technique with the three-channel image (RGB), although it indicated paths for improvement. The pencil filter also proved to be more robust than the LBP filter, as expected.

show abstract

Background Subtraction Using Local SVD Binary Pattern

Cited by 48 publications

References 23 publications

FAIM: Vision and Weight Sensing Fusion Framework for Autonomous Inventory Monitoring in Convenience Stores

FAIM: Vision and Weight Sensing Fusion Framework for Autonomous Inventory Monitoring in Convenience Stores

Data fusion of multi-spectral cameras on a low-power processing platform for self-sufficient outdoor operation

Image Processing Techniques to Improve Deep 6DoF Detection in RGB Images

Contact Info

Product

Resources

About