“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

Pattern recognition

Biofouling in marine aquaculture is a specific problem where both the target culture species and/or infrastructure are exposed to a diverse array of fouling organisms, with significant production impacts. In shellfish aquaculture the key impact is the direct fouling of stock causing physical damage, mechanical interference, biological competition and environmental modification, while infrastructure is also impacted. In contrast, the key impact in finfish aquaculture is the fouling of infrastructure which restricts water exchange, increases disease risk and causes deformation of cages and structures. Consequently, the economic costs associated with biofouling control are substantial. Conservative estimates are consistently between 5–10% of production costs (equivalent to US$ 1.5 to 3 billion yr−1), illustrating the need for effective mitigation methods and technologies. The control of biofouling in aquaculture is achieved through the avoidance of natural recruitment, physical removal and the use of antifoulants. However, the continued rise and expansion of the aquaculture industry and the increasingly stringent legislation for biocides in food production necessitates the development of innovative antifouling strategies. These must meet environmental, societal, and economic benchmarks while effectively preventing the settlement and growth of resilient multi-species consortia of biofouling organisms.

https://www.tandfonline.com/doi/pdf/10.1080/08927014.2012.700478

The impact and control of biofouling in marine aquaculture: a review

Colour is the first quality attribute of food evaluated by consumers, and is therefore an important component of food quality relevant to market acceptance. Rapid and objective measurement of food colour is required in quality control for the commercial grading of products. Computer vision is a promising technique currently investigated for food colour measurement, especially with the ability of providing a detailed characterization of colour uniformity at pixel-based level. This paper reviews the fundamentals and applications of computer vision for food colour measurement. Introduction of colour space and traditional colour measurements is also given. At last, advantages and disadvantages of computer vision for colour measurement are analyzed and its future trends are proposed.

Colour measurements by computer vision for food quality control – A review

Human populations have been concentrated along and exploiting the coastal zones for millennia. Of regions with the highest human impacts on the oceans (Halpern et al. 2008), 6 of the top 10 have recently experienced blooms or problems with jellies. I review the time lines of human population growth and their effects on the coastal environment. I explore evidence suggesting that human activities—specifically, seafood harvest, eutrophication, hard substrate additions, transport of nonindigenous species, aquaculture, and climate change—may benefit jelly populations. Direct evidence is lacking for most of these factors; however, numerous correlations show abundant jellies in areas with warm temperatures and low forage fish populations. Jelly populations fluctuate in ∼10- and ∼20-year cycles in concert with solar and climate cycles. Global warming will provide a rising baseline against which climate cycles will cause fluctuations in jelly populations. The probable acceleration of anthropogenic effects may lead...

Jellyfish and Ctenophore Blooms Coincide with Human Proliferations and Environmental Perturbations

The changes in skin and fillet color of anesthetized and exhausted Atlantic salmon were determined immediately after killing, during rigor mortis, and after ice storage for 7 d. Skin color (CIE L*, a*, b*, and related values) was determined by a Minolta Chroma Meter. Roche SalmoFan Lineal and Roche Color Card values were determined by a computer vision method and a sensory panel. Before color assessment, the stress levels of the 2 fish groups were characterized in terms of white muscle parameters (pH, rigor mortis, and core temperature). The results showed that perimortem handling stress initially significantly affected several color parameters of skin and fillets. Significant transient fillet color changes also occurred in the prerigor phase and during the development of rigor mortis. Our results suggested that fillet color was affected by postmortem glycolysis (pH drop, particularly in anesthetized fillets), then by onset and development of rigor mortis. The color change patterns during storage were different for the 2 groups of fish. The computer vision method was considered suitable for automated (online) quality control and grading of salmonid fillets according to color.

Atlantic salmon skin and fillet color changes effected by perimortem handling stress, rigor mortis, and ice storage.

Applications of imaging technologies are playing an important role in several major areas in research and industry, including process optimization, automated sorting and grading, and automated processing. This review looks at the trends in application of imaging technologies to inspection of fish and fish products – in particular reviewing the applications of VIS/NIR imaging, VIS/NIR imaging spectroscopy, planar and computed tomography (CT) X-ray imaging, and magnetic resonance imaging (MRI). In recent years, a large amount of research has been published on the use of imaging technologies for inspecting fish and fish products. This review article condenses a representative selection of recent research and industrial solutions in order to observe the general trends in how the imaging technologies are applied to inspection of fish and fish products. Based on these observed trends, we provide some viewpoints on the current situation and suggestions for future research directions.

Trends in application of imaging technologies to inspection of fish and fish products

Computer vision method was used to evaluate the color of Atlantic salmon (Salmo salar) fillets. Computer vision-based sorting of fillets according to their color was studied on 2 separate groups of salmon fillets. The images of fillets were captured using a digital camera of high resolution. Images of salmon fillets were then segmented in the regions of interest and analyzed in red, green, and blue (RGB) and CIE Lightness, redness, and yellowness (Lab) color spaces, and classified according to the Roche color card industrial standard. Comparisons of fillet color between visual evaluations were made by a panel of human inspectors, according to the Roche SalmoFan lineal standard, and the color scores generated from computer vision algorithm showed that there were no significant differences between the methods. Overall, computer vision can be used as a powerful tool to sort fillets by color in a fast and nondestructive manner. The low cost of implementing computer vision solutions creates the potential to replace manual labor in fish processing plants with automation.

/pdf/computer-vision-based-sorting-of-atlantic-salmon-salmo-salar-5v2777ahyd.pdf

Ekrem Misimi

Papers

Atlantic salmon skin and fillet color changes effected by perimortem handling stress, rigor mortis, and ice storage.

Trends in application of imaging technologies to inspection of fish and fish products

Computer Vision-Based Sorting of Atlantic Salmon (Salmo salar) Fillets According to Their Color Level

The development of biofouling, particularly the hydroid Ectopleura larynx, on commercial salmon cage nets in Mid-Norway

A spatio-temporal recurrent network for salmon feeding action recognition from underwater videos in aquaculture