29:2 Y.Yin et al. 1 INTRODUCTION 1.1 Motivation Nowadays,smart phones are widely used in our daily lives.Due to the portability of smart phones, more and more people tend to take photos with their smart phones,for example,taking photos at a tourist attraction.However,energy-saving continues to be an upsetting problem for smart camera phones,since photographing is a very power-hungry function.For example,according to KS Mobile's(KS Mobile Inc.2014)report in 2014,the application Camera360 Ultimate is listed in first place of the top 10 battery-draining applications for Android.Therefore,the huge energy consumption becomes a non-negligible pain point for the users of smart camera phones.Conse- quently,it is essential to reduce the unnecessary energy consumption during photographing to extend the battery life of smart camera phones. 1.2 Limitations of Prior Art Prior work on energy saving of smart phones can be classified into the following three parts:energy consumption of hardware(Fan et al.2007;Bellosa et al.2003;Rajan et al.2006; Balasubramanian et al.2009),power consumption models,and energy-saving schemes for specific applications.For hardware,Chen et al.(2013a)analyze the power consumption of AMOLED displays in multimedia applications and reveal that camera recording incurs high power cost. LiKamWa et al.(2013)report the experimental and analytical characterization of CMOS image sensors and reveal two energy-proportional mechanisms for energy saving.For models,Dong and Zhong(2011)propose Sesame,with which a mobile system constructs an energy model of itself without any external assistance.Xu et al.(2013)propose a new way called V-edge to generate power models based on battery voltage dynamics.For specific applications,Han et al.(2013)study the energy cost made by human-screen interaction,such as scrolling on the screen.Dietrich and Chakraborty(2013)detect the game's current state and decrease the processor's voltage and fre- quency whenever possible to save energy.Hu et al.(2013)propose a Mobility-Assisted User Contact detection algorithm(MAUC),which utilizes the accelerometer of the phone to detect user move- ments for energy-saving.The Bluetooth scans only when user movements have a high possibility to cause contact changes.LiKamWa et al.(2013)improve the energy efficiency of image sensors based on hardware modifications.There are fewer energy-saving schemes for photographing. Being different from these prior work,we aim to propose an energy-saving scheme for pho- tographing.We aim to recognize the user's activity and reduce unnecessary energy cost when the user is not taking photos.The scheme does not need hardware modifications and user interaction, to guarantee a good user experience. 1.3 Proposed Approach A straight solution to reduce energy cost is to turn off the camera or screen while not taking photos.However,frequently turning ON/OFF the camera or screen is very annoying and leads to a bad user experience.Besides,frequently turning on the camera or screen will incur high energy consumption.Take the Samsung Galaxy Nexus smart phone as an example,the energy consumption of the pair of operations,that is,turning off the camera and screen and then turning on the screen and camera,can keep the camera working on preview mode for about 7s. To propose an efficient energy-saving scheme,we conduct extensive observations.We find that during photographing.a fairly large proportion of energy is wasted in preparations before shoot- ing.For example,the user usually first turns on the camera.Then,he/she will probably adjust the phone time and again,to find a good camera view.Finally,when the camera focuses on the target, the user will press the button to shoot.Between two consecutive shots,the camera works with ACM Transactions on Sensor Networks,Vol 13.No.4,Article 29.Publication date:September 2017.29:2 Y. Yin et al. 1 INTRODUCTION 1.1 Motivation Nowadays, smart phones are widely used in our daily lives. Due to the portability of smart phones, more and more people tend to take photos with their smart phones, for example, taking photos at a tourist attraction. However, energy-saving continues to be an upsetting problem for smart camera phones, since photographing is a very power-hungry function. For example, according to KS Mobile’s (KS Mobile Inc. 2014) report in 2014, the application Camera360 Ultimate is listed in first place of the top 10 battery-draining applications for Android. Therefore, the huge energy consumption becomes a non-negligible pain point for the users of smart camera phones. Conse￾quently, it is essential to reduce the unnecessary energy consumption during photographing to extend the battery life of smart camera phones. 1.2 Limitations of Prior Art Prior work on energy saving of smart phones can be classified into the following three parts: energy consumption of hardware (Fan et al. 2007; Bellosa et al. 2003; Rajan et al. 2006; Balasubramanian et al. 2009), power consumption models, and energy-saving schemes for specific applications. For hardware, Chen et al. (2013a) analyze the power consumption of AMOLED displays in multimedia applications and reveal that camera recording incurs high power cost. LiKamWa et al. (2013) report the experimental and analytical characterization of CMOS image sensors and reveal two energy-proportional mechanisms for energy saving. For models, Dong and Zhong (2011) propose Sesame, with which a mobile system constructs an energy model of itself without any external assistance. Xu et al. (2013) propose a new way called V-edge to generate power models based on battery voltage dynamics. For specific applications, Han et al. (2013) study the energy cost made by human-screen interaction, such as scrolling on the screen. Dietrich and Chakraborty (2013) detect the game’s current state and decrease the processor’s voltage and fre￾quency whenever possible to save energy. Hu et al. (2013) propose a Mobility-Assisted User Contact detection algorithm (MAUC), which utilizes the accelerometer of the phone to detect user move￾ments for energy-saving. The Bluetooth scans only when user movements have a high possibility to cause contact changes. LiKamWa et al. (2013) improve the energy efficiency of image sensors based on hardware modifications. There are fewer energy-saving schemes for photographing. Being different from these prior work, we aim to propose an energy-saving scheme for pho￾tographing. We aim to recognize the user’s activity and reduce unnecessary energy cost when the user is not taking photos. The scheme does not need hardware modifications and user interaction, to guarantee a good user experience. 1.3 Proposed Approach A straight solution to reduce energy cost is to turn off the camera or screen while not taking photos. However, frequently turning ON/OFF the camera or screen is very annoying and leads to a bad user experience. Besides, frequently turning on the camera or screen will incur high energy consumption. Take the Samsung Galaxy Nexus smart phone as an example, the energy consumption of the pair of operations, that is, turning off the camera and screen and then turning on the screen and camera, can keep the camera working on preview mode for about 7s. To propose an efficient energy-saving scheme, we conduct extensive observations. We find that during photographing, a fairly large proportion of energy is wasted in preparations before shoot￾ing. For example, the user usually first turns on the camera. Then, he/she will probably adjust the phone time and again, to find a good camera view. Finally, when the camera focuses on the target, the user will press the button to shoot. Between two consecutive shots, the camera works with ACM Transactions on Sensor Networks, Vol. 13, No. 4, Article 29. Publication date: September 2017