Depth Image Selection Based on Posture for Calf Body Weight Estimation

We are developing a system to estimate body weight using calf depth images taken in a loose barn. For this purpose, depth images should be taken from the side, without calves overlapping and without their backs bent. However, most of the depth images that are taken successively and automatically do not satisfy these conditions. Therefore, we need to select only the depth images that match these conditions, as to take many images as possible. The existing method assumes that a calf standing sideways and upright in front of cameras is in a suitable pose. However, since such cases rarely occur, not many images were selected. This paper proposes a new depth image-selection method, focusing on whether a calf is sideways, and the back is not bent, regardless of whether the calf is still or walking. First, depth images including only a single calf are extracted. The calf was identified using radio frequency identification (RFID) when its depth image was taken. Then, the calf area was extracted by background subtraction and contour detection with a depth image. Finally, to judge the usable depth images, we detected and evaluated the calf’s posture, such as the angle of the calf to the camera and the slope of the dorsal line. We used the mean absolute percentage error (MAPE) to assess the efficiency of our method. As two times the number of depth images were extracted, our method achieved an MAPE of 12.45%, while the existing method achieved an MAPE of 13.87%. From this result, we have confirmed that our method makes body weight estimation more accurate.

Download Full-text

Repeatability and frequency of in-paddock sheep walk-over weights: implications for individual animal management

Animal Production Science ◽

10.1071/an12311 ◽

2014 ◽

Vol 54 (2) ◽

pp. 207 ◽

Cited By ~ 8

Author(s):

D. J. Brown ◽

D. B. Savage ◽

G. N. Hinch

Keyword(s):

Radio Frequency Identification ◽

Null Hypothesis ◽

Average Reference ◽

Animal Management ◽

Data Points ◽

Frequency Identification ◽

The Mean ◽

Gut Fill ◽

On Farm ◽

Over Time

Sheep liveweight is an indicator of nutritional status, and its measure may be used as an aid to nutritional management. When walk-over weighing (WOW), a remote weighing concept for grazing sheep, is combined with radio frequency identification (RFID), resulting ‘RFID-linked WOW’ data may enable the liveweight of individual sheep to be tracked over time. We investigated whether RFID-linked WOW data is sufficiently repeatable and frequent to generate individual liveweight estimates with 95% confidence intervals (95% CI) of <2 kg (a sufficient level of error to account for fluctuating gut fill) for a flock within timeframes suitable for management (1-day and 5-day timeframes). Four flocks of sheep were used to generate RFID-linked WOW datasets. RFID-linked WOW data were organised into three groups: raw (unfiltered), coarse filtered (remove all sheep-weights outside the flock’s liveweight range), and fine filtered (remove all sheep-weights outside a 25% range of a recent flock average reference liveweight). The repeatability of raw (unfiltered) RFID-linked WOW data was low (0.20), while a coarse (0.46) and fine (0.76) data filter improved repeatability. The 95% CI of raw RFID-linked WOW data was 27 kg, and was decreased by a coarse (11 kg) and fine (6 kg) data filter. Increasing the number of raw, coarse and fine-filtered data points to 190, 30 and 12 sheep-weights, respectively, decreased the 95% CI to <2 kg. The mean cumulative percentage of sheep achieving >11 fine-filtered RFID-linked WOW sheep-weights within a 1-day and 5-day timeframe was 0 and 10%, respectively. The null hypothesis was accepted: RFID-linked WOW data had low repeatability and was unable to generate liveweight estimates with a 95% CI of less than 2 kg within a suitable timeframe. Therefore, at this stage, RFID-linked WOW is not recommended for on-farm decision making of individual sheep.

Download Full-text

Estimation of fetal weight by ultrasonography after preterm premature rupture of membranes: comparison of different formulas

Journal of Perinatal Medicine ◽

10.1515/jpm-2016-0258 ◽

2017 ◽

Vol 45 (2) ◽

Cited By ~ 2

Author(s):

Sertac Esin ◽

Mutlu Hayran ◽

Yusuf Aytac Tohma ◽

Mahmut Guden ◽

Ismail Alay ◽

...

Keyword(s):

Birth Weight ◽

Gestational Age ◽

Fetal Weight ◽

Percentage Error ◽

Weight Estimation ◽

Premature Rupture ◽

Preterm Premature Rupture ◽

Absolute Percentage Error ◽

The Mean ◽

Weight Groups

AbstractObjective:To compare different ultrasonographic fetal weight estimation formulas in predicting the fetal birth weight of preterm premature rupture of membrane (PPROM) fetuses.Methods:Based on the ultrasonographic measurements, the estimated fetal weight (EFW) was calculated according to the published formulas. The comparisons used estimated birth weight (EBW) and observed birth weight (OBW) to calculate the mean absolute percentage error [(EBW–OBW)/OBW×100], mean percentage error [(EBW–OBW)/OBW×100)] and their 95% confidence intervals.Results:There were 234 PPROM patients in the study period. The mean gestational age at which PPROM occured was 31.2±3.7 weeks and the mean gestational age of delivery was 32.4±3.2 weeks. The mean birth weight was 1892±610 g. The median absolute percentage error for 33 formulas was 11.7%. 87.9% and 21.2% of the formulas yielded inaccurate results when the cut-off values for median absolute percentage error were 10% and 15%, respectively. The Vintzileos’ formula was the only method which had less than or equal to 10% absolute percentage error in all age and weight groups.Conclusions:For PPROM patients, most of the formulas designed for sonographic fetal weight estimation had acceptable performance. The Vintzileos’ method was the only formula having less than 10% absolute percentage error in all gestational age and weight groups; therefore, it may be the preferred method in this cohort. Amniotic fluid index (AFI) before delivery had no impact on the performance of the formulas in terms of mean percentage errors.

Download Full-text

58 RADIO-FREQUENCY IDENTIFICATION INTELI-STRAWS: GAMETES AND EMBRYO PACKAGING, STORAGE, AND INFORMATION RECOVERY

Reproduction Fertility and Development ◽

10.1071/rdv29n1ab58 ◽

2017 ◽

Vol 29 (1) ◽

pp. 136

Author(s):

M. E. Kjelland ◽

T. Loper ◽

C. Woodley ◽

T. M. Swannack ◽

T. K. Stroud ◽

...

Keyword(s):

Radio Frequency ◽

Liquid Nitrogen ◽

Radio Frequency Identification ◽

Exact Test ◽

Information Recovery ◽

Cold Tolerant ◽

Frequency Identification ◽

The Mean ◽

Extreme Cold ◽

And Control

The assisted reproduction industry involving sales and services for gametes and embryos for domestic animals of commercial value is a large market totaling millions of dollars annually. The objective of this study was to develop and test gamete and embryo packaging—Inteli-Straws (I-S) equipped with radio-frequency identification (RFID) technology. Specifically, French straws (0.25 and 0.5 mL) were modified to include extreme cold-tolerant RFID microchips. Two groups of I-S were formed: Group (G)1: RFID chips that were autoclaved (n = 49), and G2: RFID chips that were not autoclaved (n = 47). Both groups had a control that was not exposed to liquid nitrogen (LN). Each group was exposed to LN up to 4 times: 2 slow freezes first and then 2 fast (i.e. vitrification) freezes, and I-S RFID chip survival was determined. I-S detection and readability (non-autoclaved) was also measured, placing I-S just above LN (in vapors, n = 43) or just below LN (n = 38). Statistical differences (α = 0.05) were determined using Fisher’s exact test. The results between G1 and G2 were not significantly different (P = 0.108) after 4 rounds of cryopreservation (and thawing). For G1, 98% (48/49) of the I-S RFID chips remained operational, and control and treatment were not significantly different (P = 1.000). For G2, 89.4% (42/47) of the autoclaved RFID chips remained operational, and control and treatment not significantly different (P = 0.099). RFID chip readability results; that is, the ability to detect the I-S versus not able to detect the I-S, comparing placement just above liquid nitrogen (LN) versus the placement just below LN were not significantly different (P = 0.105). Notably, detection differences varied within each group, with I-S in G1 (mean = 9.5; SD = 3.5 cm) readable at a larger distance, 5.2 cm farther than the mean of G2 (mean = 4.3; SD = 1.9 cm). During AI or embryo transfer (ET), a technician may not clearly identify the label or colour of straw, may incorrectly record the information, or may take more time than desirable to record it. Increased exposure times may lead to decreased viability of gametes and embryos. The results show that by using the I-S, one may quickly scan the straw within LN or LN vapors, thereby automatically detecting information and even uploading it to a database (e.g. scanner sophistication). We are not aware of comparable device to I-S for locating and retrieving associated information without removing the gamete/embryo packaging from LN or LN vapors; unlike traditionally labelled straws (e.g. laser etched or ink labels). Also, for AI and ET, the I-S can be quickly scanned and the straw information automatically detected and uploaded to a database.

Download Full-text

Automated monitoring of mouse feeding and body weight for continuous health assessment

Laboratory Animals ◽

10.1177/0023677218797974 ◽

2018 ◽

Vol 53 (4) ◽

pp. 342-351 ◽

Cited By ~ 3

Author(s):

Jamie Ahloy-Dallaire ◽

Jon D Klein ◽

Jerry K Davis ◽

Joseph P Garner

Keyword(s):

Body Weight ◽

Feed Intake ◽

Radio Frequency Identification ◽

Home Cage ◽

Objective Assessment ◽

Health Assessment ◽

Automated Monitoring ◽

Laboratory Rodents ◽

Significant Difference ◽

Frequency Identification

Routine health assessment of laboratory rodents can be improved using automated home cage monitoring. Continuous, non-stressful, objective assessment of rodents unaware that they are being watched, including during their active dark period, reveals behavioural and physiological changes otherwise invisible to human caretakers. We developed an automated feeder that tracks feed intake, body weight, and physical appearance of individual radio frequency identification-tagged mice in social home cages. Here, we experimentally induce illness via lipopolysaccharide challenge and show that this automated tracking apparatus reveals sickness behaviour (reduced food intake) as early as 2–4 hours after lipopolysaccharide injection, whereas human observers conducting routine health checks fail to detect a significant difference between sick mice and saline-injected controls. Continuous automated monitoring additionally reveals pronounced circadian rhythms in both feed intake and body weight. Automated home cage monitoring is a non-invasive, reliable mode of health surveillance allowing caretakers to more efficiently detect and respond to early signs of illness in laboratory rodent populations.

Download Full-text

Frequent Visits to an Outdoor Range and Lower Areas of an Aviary System Is Related to Curiosity in Commercial Free-Range Laying Hens

Animals ◽

10.3390/ani10091706 ◽

2020 ◽

Vol 10 (9) ◽

pp. 1706

Author(s):

Manisha Kolakshyapati ◽

Peta Simone Taylor ◽

Adam Hamlin ◽

Terence Zimazile Sibanda ◽

Jessica de Souza Vilela ◽

...

Keyword(s):

Body Weight ◽

Radio Frequency ◽

Radio Frequency Identification ◽

Early Life ◽

Life Time ◽

Novel Object ◽

Free Range ◽

Lower Tier ◽

Nest Boxes ◽

Frequency Identification

Individual hen preferences to spend time at particular locations within a free-range aviary system and relationships with temperament is relatively unknown. Hens (n = 769) from three commercial flocks were monitored with Radio Frequency Identification technology to determine time spent on the range, upper and lower aviary tiers, and nest boxes. Prior depopulation, novel arena (NA) and novel object (NO) tests assessed exploration and fearfulness. During early life; more time on the lower tier was associated with more lines crossed in the NA test (p < 0.05). No other evidence suggested preference during early life was related to fear or curiosity. More time on the range and lower tier were associated with heavier pre-ranging body weight and gain (p = 0.0001). Over the hens’ whole life; time spent on range and lower tier was associated with approaching the NO (p < 0.01). More time spent on the upper tier was associated with less time near the NO and fewer lines crossed in NA (p < 0.01). The relationships during early and whole life use of space and some potential indicators of fearfulness were inconsistent and therefore, no strong, valid, and reliable indicators of hen fearfulness such as freezing were identified.

Download Full-text

Anomaly Segmentation Based on Depth Image for Quality Inspection Processes in Tire Manufacturing

Applied Sciences ◽

10.3390/app112110376 ◽

2021 ◽

Vol 11 (21) ◽

pp. 10376

Author(s):

Dongbeom Ko ◽

Sungjoo Kang ◽

Hyunsuk Kim ◽

Wongok Lee ◽

Yousuk Bae ◽

...

Keyword(s):

Deep Learning ◽

Manufacturing Industry ◽

Industrial Revolution ◽

Depth Image ◽

Learning Technology ◽

Limited Information ◽

Depth Images ◽

General Training ◽

The Mean ◽

Deep Learning Network

This paper introduces and implements an efficient training method for deep learning–based anomaly area detection in the depth image of a tire. A depth image of 16 bit integer size is used in various fields, such as manufacturing, industry, and medicine. In addition, the advent of the 4th Industrial Revolution and the development of deep learning require deep learning–based problem solving in various fields. Accordingly, various research efforts use deep learning technology to detect errors, such as product defects and diseases, in depth images. However, a depth image expressed in grayscale has limited information, compared with a three-channel image with potential colors, shapes, and brightness. In addition, in the case of tires, despite the same defect, they often have different sizes and shapes, making it difficult to train deep learning. Therefore, in this paper, the four-step process of (1) image input, (2) highlight image generation, (3) image stacking, and (4) image training is applied to a deep learning segmentation model that can detect atypical defect data. Defect detection aims to detect vent spews that occur during tire manufacturing. We compare the training results of applying the process proposed in this paper and the general training result for experiment and evaluation. For evaluation, we use intersection of union (IoU), which compares the pixel area where the actual error is located in the depth image and the pixel area of the error inferred by the deep learning network. The results of the experiment confirmed that the proposed methodology improved the mean IoU by more than 7% and the IoU for the vent spew error by more than 10%, compared to the general method. In addition, the time it takes for the mean IoU to remain stable at 60% is reduced by 80%. The experiments and results prove that the methodology proposed in this paper can train efficiently without losing the information of the original depth data.

Download Full-text