Personal Healthy Diet and Calorie Monitoring System using Fuzzy Inference in Smart Phones

應用於智慧型手機的個人健康飲食與熱量監控系統使用模糊推論

秦群立(Chiun-Li Chin)；林若真(Ruo-Jhen Lin)；劉士華(Shih-Hua Liu)

智慧型手機 ； 熱量 ； 模糊推論 ； 重力感測裝置 ； Smart phones ； Calories ； QR code ； Fuzzy inference ； G-Sensor

中山醫學雜誌

24卷1期（2013 / 06 / 01）

49 - 60

英文

由於現代外食族的普及、肥胖人口的增加、成年人運動量少且加上飲食精緻化，造成飲食不均衡面臨健康危機，使得減重成為目前國人重要關注的議題之一，因此開始越來越關心熱量及營養成份的攝取，開始重視自己所攝取的食物是否健康及均衡營養，所以我們開發出一套「應用於智慧型手機的個人健康飲食與熱量監控系統使用模糊推論」。我們利用智慧型手機的可攜性、方便性與不受距離限制的特性，並且利用QR code的便利性和資料儲存量大等特點來開發出一套結合模糊推論（Fuzzy inference）技術之系統，系統會讀取QR code中的營養成份，包含食物的熱量、蛋白質、脂肪、碳水化合物和鈉，接著利用手機中的重力感測裝置（G-Sensor）來測量使用者運動時的搖晃程度，算出使用者運動時消耗的熱量，最後計算個人攝取營養成份的總量，藉以判斷是否有超出標準量。實驗結果的部份，本系統成功率90%，本系統與Nike+ sensor做比較，所測量出來的消耗熱量，以本系統和標準消耗熱量公式較為相近。所以不論是想瘦身的男性或女性、必須注意飲食的慢性病患者、發育中的孩童或者是需要均衡飲食的壯年人，有一套可以讓他們輕鬆選擇食物的系統，如此可針對不同的族群調整適合自己的營養吸收量，同時還能利用本系統來運動消耗熱量，達到均衡營養與熱量控制的目的。

The increasing popularity of dining out has led to a rise in obesity and a reduction in exercise among adults. In addition, the increasing consumption of refined foods is resulting in unbalanced diets and elevated risk of various chronic diseases. To reduce these risks, much attention is being paid to methods of weight loss in Taiwan, and there is increasing awareness of the need to monitor caloric and nutrient intake, as well as an emphasis on the healthiness and balanced nature of food intake. Hence, we developed the ”personal diet and caloric intake monitoring system” for use on smart phones, which is based on the fuzzy inference method. In this system, a QR code is used to save information regarding nutrients, including calories, protein, fat, carbohydrates, and sodium. Using the data stored in the QR code, the proposed system calculates the total nutrition from individual sessions of food intake. Next, it determines whether the number of intake-calories and intake-nutrition value go beyond the daily recommended allowances. In addition, this system applies the G-Sensor available on most smart phones to measure the rocking motion of users during exercise. Fuzzy inference method infers number of calories burned. The success rate of the system was 90% in terms of the accuracy of calorie burning calculations. Based on comparisons with Nike+ sensor, the calorie-burning value measured by the proposed system was closer to that calculated by the standard calorie-burning formula. In summary, this proposed system was able to help users make food selections and monitor calories, regardless of gender or goal whether it be losing weight or monitoring diet. Users are able to adjust nutritional intake and measure calories burned during exercise to eventually balance nutrients and control caloric intake.