Mapeamento projetivo, associação de palavras e escala hedônica como instrumentos de avaliação da percepção de produtos de pescado por crianças da rede pública de ensino

Brazil is a country that is characterized by its low consumption of fish. With consumption records of 10.6 kg/ inhabitant/ year, it is lower than the recommended by the UN, that is 12 kg/ inhabitant/ year. The regular consumption of fish provides health gain for people and their introduction into the school feeding is an important strategy for the insertion of this food consumption habits in a population. In this context, the objective of this study was to understand the perception of fish with children from the public school system through the technical Projective Mapping (MP) and Association of Words (AP); and evaluate the acceptability of fish derivative in school meals. In the first instance with the intention to better understand the perception of children from different ages about the fish-based products, Projective Mapping techniques were applied through the use of food figures and word association. A total of 149 children from three public schools from Pato Branco, Paraná State, Brazil, took part in this study. Three groups of children aged 5-6, 7-8 and 9-10 years old were interviewed individually by six monitors experienced in applied sensory methods. Ten figures with healthy foods drawings (sushi, salad, fruit, fish, chicken), and less healthy foods (pizza, pudding, cake, hamburger, fries) were distributed to the children, who were asked to paste the figures in A3 sheet, so that the products they considered similar stayed near each other, and the ones considered very different stayed apart. After this, the children described the images and the image groups (Ultra Flash Profile). The results revealed that the MP technique was easily operated and understood by all the children and the use of images made its implementation easier. The results analysis also revealed different perceptions came from children from different ages and hedonic perceptions regarding the fish-based products had a greater weight in the percentage from older children. AP technique proved to be an important tool to understand the perception of fish by children, and strengthened the results previously obtained by the MP. In a second step it was evaluated the acceptance of fish burger (tilapia) in school meals. For this task, the school cooks were trained to prepare the hamburgers. For the evaluation of acceptance, the hedonic scale was used with 5 facial ratings (1 = disliked very much to 5 = liked a lot). Students from both genders, between 5 to 10 years old (n = 142) proved the burgers at lunchtime, representing the protein portion of the meal. The tilapia derivative products shown to be foods with important nutritional value and low calorie value. For the application of the multinomial logistic regression analysis there was no significant effect from the age and gender variation in the acceptance by children. However, statistical significance was determined in the interaction between these two variables. With 87 % acceptance rate there was potential for consumption of fish burgers in school meals.

A contribution to semantic description of images and videos: an application of soft biometrics

Humans have a high ability to extract visual data information acquired by sight. Trought a learning process, which starts at birth and continues throughout life, image interpretation becomes almost instinctively. At a glance, one can easily describe a scene with reasonable precision, naming its main components. Usually, this is done by extracting low-level features such as edges, shapes and textures, and associanting them to high level meanings. In this way, a semantic description of the scene is done. An example of this, is the human capacity to recognize and describe other people physical and behavioral characteristics, or biometrics. Soft-biometrics also represents inherent characteristics of human body and behaviour, but do not allow unique person identification. Computer vision area aims to develop methods capable of performing visual interpretation with performance similar to humans. This thesis aims to propose computer vison methods which allows high level information extraction from images in the form of soft biometrics. This problem is approached in two ways, unsupervised and supervised learning methods. The first seeks to group images via an automatic feature extraction learning , using both convolution techniques, evolutionary computing and clustering. In this approach employed images contains faces and people. Second approach employs convolutional neural networks, which have the ability to operate on raw images, learning both feature extraction and classification processes. Here, images are classified according to gender and clothes, divided into upper and lower parts of human body. First approach, when tested with different image datasets obtained an accuracy of approximately 80% for faces and non-faces and 70% for people and non-person. The second tested using images and videos, obtained an accuracy of about 70% for gender, 80% to the upper clothes and 90% to lower clothes. The results of these case studies, show that proposed methods are promising, allowing the realization of automatic high level information image annotation. This opens possibilities for development of applications in diverse areas such as content-based image and video search and automatica video survaillance, reducing human effort in the task of manual annotation and monitoring.