Search

The average weight and flock uniformity of broilers in group housing is important information that allows producers to know the flock growth conditions and determine the selling time. However, gathering weight information of chickens is not only labor-intensive for humans but also frightening for the birds. In this study, an image-assisted rod-platform weighing system was developed to automatically monitor the average weight and flock uniformity of broilers in chicken houses. This weighing system consists of a computer and several weighing scales. Each weighing scale contains a rod-platform weighing module and a surveillance camera module. The principle of the automated weighing system is to estimate population weight information using samples. The design of the rod-platform weighing module was based on the perching habit of birds to attract more broilers to stand on the rod platform and thus get more weight samples. The surveillance camera module is used to detect the number of broilers on the rod using image processing technology. A data processing method called PORWI, which includes elimination of redundant records and trim of non-redundant records, was designed to optimize the results of chicken number identification from images to improve the accuracy of the results. An experiment was done in two small groups of broilers with approximately 100 chickens and 8.58 m2 of area for each group. A weekly weighing was conducted, and three kinds of weight information were obtained, which included manual population weight information (MPWI), manual sample-based weight information (MSWI), and automated sampling weight information (ASWI). Each weight information set comprised the group average weight and flock uniformity, which were then used to evaluate accuracy. The perching rate of chickens using the rod platform reached an average of 60 times h-1, and the rate was retained with increasing age. Compared with the MPWI obtained by individual weighing, the manual sample-based measurement method provided results with errors of 0% to +5%, while our automated weighing system achieved accuracies within ±2% for average weight and ±1.5% for flock uniformity. Keywords: Automatic weighing, Average weight, Broiler, Chicken detection, Uniformity.

Automatic detection of fruit peel defects by a computer vision system is difficult due to the challenges of acquiring images from the surface of spherical fruit and the visual similarity between the stem-ends and the true defects. In this study, oranges with wind scarring, thrips scarring, scale infestation, dehiscent fruit, anthracnose, copper burn, canker spot and normal surface were researched. A lighting transform method based on a low pass Butterworth filter with a cutoff frequency D 0 = 7 was first developed to convert the non-uniform intensity distribution on spherical oranges into a uniform intensity distribution over the whole fruit surface. However, the stem-ends were easily confused with defective areas. In order to solve this problem, different color components (R, G and B) and their combinations were analyzed. It was found that a ratio method and R and G component combination coupled with a big area and elongated region removal algorithm (BER) could be used to differentiate stem-ends from defects effectively. Finally, a processing and classification algorithm based on a simple thresholding method was proposed. The result with 98.9% overall detection rate for the 720 independent sample images indicated that the proposed algorithm was effective in differentiation of normal and defective oranges. The method, however, could not discriminate between different types of defects.

Shape is one of the most important indexes for fruit quality evaluation, and extracting the shape characters of fruit is the key step for the shape grading. However, it is very difficult to get the accurate description of fruit shape because of a variety of fruit shape in different attitudes. In this research, 4 images were taken for each apple respectively. There are 144 images for 36 apples, in which 12 apples being considered to have the best shape, and 11 ones having big deformation, and the other 13 ones having small deformation. Then process these images and get 245 shape character parameters of each boundary of the fruit through signature description, Hu moments, Zernike moments, wavelet moments, Fourier description, and wavelet description. This paper is trying to get efficient and practical shape indexes and find a feasible way for describing fruit shape.