Region-Based CNN for Anomaly Detection in PV Power Plants Using Aerial Imagery

Vlaminck, Michiel; Heidbuchel, Rugen; Philips, Wilfried; Luong, Hiêp

doi:10.3390/s22031244

Cited by 31 publications

(14 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, there is an increasing awareness as to how modern computer vision techniques, and deep learning in particular, can be applied to satellite imagery to tackle environmental and regulatory challenges [22]. Our paper also bears some relationship to previous work in anomaly detection in aerial imagery, such as detecting failures in power plants [59] and building damage detection [56]. Indeed, it is straightforward to cast our problem in terms of anomaly detection.…”

Section: Related Workmentioning

confidence: 94%

Detecting Environmental Violations with Satellite Imagery in Near Real Time: Land Application under the Clean Water Act

Chugg,

Rothbacher,

Feng

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper introduces a new, highly consequential setting for the use of computer vision for environmental sustainability. Concentrated Animal Feeding Operations (CAFOs) (aka intensive livestock farms or "factory farms") produce significant manure and pollution. Dumping manure in the winter months poses significant environmental risks and violates environmental law in many states. Yet the federal Environmental Protection Agency (EPA) and state agencies have relied primarily on self-reporting to monitor such instances of "land application." Our paper makes four contributions. First, we introduce the environmental, policy, and agricultural setting of CAFOs and land application. Second, we provide a new dataset of high-cadence (daily to weekly) 3m/pixel satellite imagery from 2018-20 for 330 CAFOs in Wisconsin with hand labeled instances of land application (n=57,697). Third, we develop an object detection model to predict land application and a system to perform inference in near real-time. We show that this system effectively appears to detect land application (PR AUC = 0.93) and we uncover several outlier facilities which appear to apply regularly and excessively. Last, we estimate the population prevalence of land application events in Winter 2021/22. We show that the prevalence of land application is much higher than what is self-reported by facilities. The system can be used by environmental regulators and interest groups, one of which piloted field visits based on this system this past winter. Overall, our application demonstrates the potential for AI-based computer vision systems to solve major problems in environmental compliance with near-daily imagery.

show abstract

Section: Related Workmentioning

confidence: 94%

Detecting Environmental Violations with Satellite Imagery in Near Real Time: Land Application under the Clean Water Act

Chugg,

Rothbacher,

Feng

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Due to their reliance on classic image processing all these techniques require extensive manual hyper parameter tuning and do not generalize well to unseen imagery. Furthermore, these works only detect the <10/3/1 -Binary thresholding [110] 4/4/1 -k-means clustering of pixel intensities [111] 120/120/1 -Fuzzy rule-based classification [112] 315/315/1 -Color image descriptors [113] <10/2/1 -Iterative growth of segmentation mask [114] 375/375/1 -HoG features + naïve Bayes classifier [79] <10/-/1 -VGG-16 semantic segmentation [115] 9000/9000/6 -Faster R-CNN object detection [116] 783/783/1 -Supervised classification with CNN [117] 1000/1000/>1 -Supervised classification with CNN [118] n.a.…”

Section: Detection Of Pv Module Anomaliesmentioning

confidence: 99%

Infrared imaging of photovoltaic modules: a review of the state of the art and future challenges facing gigawatt photovoltaic power stations

Buerhop‐Lutz

Bommes

Schlipf³

et al. 2022

Prog. Energy

View full text Add to dashboard Cite

Thermography is a frequently used and appreciated method to detect underperforming PV modules in PV power stations. With the review, we give insights on two aspects: 1) are the developed measurement strategies highly efficient (about 1 module per second) to derive timely answers from the images for operators of multi-MWp power stations, and 2) do PV stakeholders get answers on the relevance of thermal anomalies for further decisions. Following these questions, the influence of measurement conditions, image and data collection, image evaluation as well as image assessment are discussed. From the literature it is clear that automated image acquisition with manned and unmanned aircrafts allow to capture more than 1 module per second. This makes it possible to achieve almost identical measurement conditions for the modules; however, it is documented to what extent the increase in speed is achieved at the expense of image resolution. Many image processing tools based on machine learning have been developed and show the potential for analysis of IR images and defect classification. There are different approaches to evaluating IR anomalies in terms of impact on performance, yield or degradation, of individual modules or modules in a string configuration. It is clear that the problem is very complex and multi-layered. On the one hand, information on the electrical interconnection is necessary, and on the other hand, there is a lack of sufficient and suitable data sets to adapt existing computer vision tools to PV. This is where we see the greatest need for action and further development to increase the expressiveness of IR images for PV stakeholder. We conclude with recommendations to improve the outcome of IR-images and encourage the generation of suitable public data sets of IR-footage for the development of machine learning tools.

show abstract

“…The automatic optical inspection of large-scale equipment surfaces, such as solar plantations, using aerial imagery is demonstrably highly effective. Vlaminck et al [11] proposed a two-stage approach to the automatic detection of anomalies in large photovoltaic sites using drone-based imaging. Similarly, Tommaso et al [12] presented a UAV-based inspection system for improved photovoltaic diagnostics based on a multistage architecture built on top of YOLOv3 [10].…”

Section: Introductionmentioning

confidence: 99%

Automated optical inspection of FAST’s reflector surface using drones and computer vision

Li¹,

Jiang²,

Song³

et al. 2023

gxjzz

View full text Add to dashboard Cite

The Five-hundred-meter Aperture Spherical radio Telescope (FAST) is the world's largest single-dish radio telescope. Its large reflecting surface achieves unprecedented sensitivity but is prone to damage, such as dents and holes, caused by naturally-occurring falling objects. Hence, the timely and accurate detection of surface defects is crucial for FAST's stable operation. Conventional manual inspection involves human inspectors climbing up and examining the large surface visually, a time-consuming and potentially unreliable process. To accelerate the inspection process and increase its accuracy, this work makes the first step towards automating the inspection of FAST by integrating deep-learning techniques with drone technology. First, a drone flies over the surface along a predetermined route. Since surface defects significantly vary in scale and show high inter-class similarity, directly applying existing deep detectors to detect defects on the drone imagery is highly prone to missing and misidentifying defects. As a remedy, we introduce cross-fusion, a dedicated plug-in operation for deep detectors that enables the adaptive fusion of multi-level features in a point-wise selective fashion, depending on local defect patterns. Consequently, strong semantics and fine-grained details are dynamically fused at different positions to support the accurate detection of defects of various scales and types. Our AI-powered drone-based automated inspection is time-efficient, reliable, and has good accessibility, which guarantees the long-term and stable operation of FAST.

show abstract

Region-Based CNN for Anomaly Detection in PV Power Plants Using Aerial Imagery

Cited by 31 publications

References 22 publications

Detecting Environmental Violations with Satellite Imagery in Near Real Time: Land Application under the Clean Water Act

Detecting Environmental Violations with Satellite Imagery in Near Real Time: Land Application under the Clean Water Act

Infrared imaging of photovoltaic modules: a review of the state of the art and future challenges facing gigawatt photovoltaic power stations

Automated optical inspection of FAST’s reflector surface using drones and computer vision

Contact Info

Product

Resources

About