Walnut Ripeness Detection Based on Coupling Information and Lightweight YOLOv4

To realize rapid and accurate ripeness detection for walnut on mobile terminals such as mobile phones, we propose a method based on coupling information and lightweight YOLOv4. First, we collected 50 walnuts at each ripeness (Unripe, Mid-ripe, Ripe, Over-ripe) to determine the kernel oil content. Pearson correlation analysis and one-way analysis of variance (ANOVA) prove that the division of walnut ripeness reflects the change in kernel oil content. It is feasible to estimate the kernel oil content by detecting the ripeness of walnut. Next, we achieve ripeness detection based on lightweight YOLOv4. We adopt MobileNetV3 as the backbone feature extractor and adopt depthwise separable convolution to replace the traditional convolution. We design a parallel convolution structure with depthwise convolution stacking (PCSDCS) to reduce parameters and improve feature extraction ability. To enhance the model’s detection ability for walnuts in the growth-intensive areas, we design a Gaussian Soft DIoU non-maximum suppression (GSDIoU-NMS) algorithm. The dataset used for model optimization contains 3600 images, of which 2880 images in the training set, 320 images in the validation set, and 400 images in the test set. We adopt a multi-training strategy based on dynamic learning rate and transfer learning to get training weights. The lightweight YOLOv4 model achieves 94.05%, 90.72%, 88.30%, 76.92 FPS, and 38.14 MB in mean average precision, precision, recall, average detection speed, and weight capacity, respectively. Compared with the Faster R-CNN model, EfficientDet-D1 model, YOLOv3 model, and YOLOv4 model, the lightweight YOLOv4 model improves 8.77%, 4.84%, 5.43%, and 0.06% in mean average precision, 74.60 FPS, 55.60 FPS, 38.83 FPS, and 46.63 FPS in detection speed, respectively. And the lightweight YOLOv4 is 84.4% smaller than the original YOLOv4 model in terms of weight capacity. This paper provides a theoretical reference for the rapid ripeness detection of walnut and exploration for the model’s lightweight.

The rheology and foamability of crystal-melt suspensions composed of triacylglycerols

The rheology of triacylglycerol (TAG) crystal-melt suspensions (CMSs) consisting of anhydrous milk fat (AMF), cocoa butter (CB), and palm kernel oil (PKO) as function of crystallization shear rate γcryst and...

Comparison of Microwave-Assisted Extraction to Soxhlet Extraction of Mango Seed Kernel Oil using Ethanol and n-Hexane as Solvents

Mango seed kernel oil was extracted by Soxhlet Extraction (SE) and Microwave-Assisted Extraction (MAE) with ethanol and n-hexane as extraction solvents. To optimize the extraction condition for SE, the temperature was set to 90°C for ethanol and 80°C for n-hexane with varying solvent-to-feed ratios (S/F ratio) of 75/12, 75/10, and 60/6 mL/g. As for MAE, the same S/F ratios were considered. Extraction was done for 5, 10, and 15 minutes with microwave power levels of 120 and 240 W. It was found out that the highest yield per extraction process for SE was: 18.00±0.25 % and 9.38±2.03 % using ethanol and n-hexane, respectively; and 6.69±0.05 % and 4.68±0.06 %using ethanol and n-hexane, respectively for MAE. It was also noted that MAE, with the microwave power level of 120 W has less extraction time for about 15 minutes as compared to SE of 8 hours. Also, the best S/F ratio in this study is 60/6 for all processes. In oil quality determination, the oil extracted was examined through several tests such as FTIR, GC-MS, acid value, % FFA, iodine value, saponification value, and melting point. It was noted that oil extracted in ethanol has a better yield compared to that of n-hexane but the oil extracted using n-hexane would provide superior quality.

SUBSTITUTION OF FISH OIL WITH PALM KERNEL OIL IN DIETS OF Oreochromis niloticus FRY: EFFECTS ON GROWTH, FEED UTILIZATION AND ECONOMIC ESTIMATES

This study assessed the impact of replacing fish oil with palm kernel oil (PKO) in the diet of Oreochromis niloticus fry on growth, feed efficiency and proximate composition. Three isonitrogenous (30% crude protein) and isolipidic (10% crude lipid) diets were formulated using palm kernel oil as a substitute for fish oil at 0% (PKO-0), 50% (PKO-50), and 100% (PKO-100). Two hundred O. niloticus fry with initial weight of 0.80± 0.25g were purchased from Water Research Institute Upper West, Ghana and transported to the Spanish Laboratory of University for Development Studies (Ghana) where they were kept and fed two times a day on commercial diet from Ranaan feed for two weeks. The fry was then stocked in triplicate groups in 60 L tanks (50 cm x 40 cm x 40 cm) at 20 fry per tank. At the end of the eight weeks feeding trial, there was a significant difference amongst the three treatments with respect to final weight, weight gain, feed conversion ratio, and specific growth rate. It was observed that the least mean values for feed intake, protein productive value, protein efficiency ratio, and protein intake occurred in fish fed PKO-0. There was a trend of increasing whole body moisture content as palm kernel oil increased. Fish fed PKO-0 recorded the lowest lipid content (7.48 ± 1.13%) in the whole body. From the economic analysis, it is evident that palm kernel oil is a cheaper source of lipid for tilapia. This study therefore recommends palm kernel oil as a substitute for tilapia diet.